paperclip/packages/plugins/sdk/src/testing.ts

2213 lines
91 KiB
TypeScript
Raw Normal View History

2026-03-13 16:22:34 -05:00
import { randomUUID } from "node:crypto";
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
import { pluginOperationIssueOriginKind } from "@paperclipai/shared";
2026-03-13 16:22:34 -05:00
import type {
PaperclipPluginManifestV1,
PluginCapability,
PluginEventType,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
PluginIssueOriginKind,
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
PluginManagedAgentResolution,
PluginManagedRoutineResolution,
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
PluginManagedSkillResolution,
CompanySkill,
2026-03-13 16:22:34 -05:00
Company,
Project,
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
Routine,
RoutineRun,
2026-03-13 16:22:34 -05:00
Issue,
IssueComment,
[codex] Add structured issue-thread interactions (#4244) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - Operators supervise that work through issues, comments, approvals, and the board UI. > - Some agent proposals need structured board/user decisions, not hidden markdown conventions or heavyweight governed approvals. > - Issue-thread interactions already provide a natural thread-native surface for proposed tasks and questions. > - This pull request extends that surface with request confirmations, richer interaction cards, and agent/plugin/MCP helpers. > - The benefit is that plan approvals and yes/no decisions become explicit, auditable, and resumable without losing the single-issue workflow. ## What Changed - Added persisted issue-thread interactions for suggested tasks, structured questions, and request confirmations. - Added board UI cards for interaction review, selection, question answers, and accept/reject confirmation flows. - Added MCP and plugin SDK helpers for creating interaction cards from agents/plugins. - Updated agent wake instructions, onboarding assets, Paperclip skill docs, and public docs to prefer structured confirmations for issue-scoped decisions. - Rebased the branch onto `public-gh/master` and renumbered branch migrations to `0063` and `0064`; the idempotency migration uses `ADD COLUMN IF NOT EXISTS` for old branch users. ## Verification - `git diff --check public-gh/master..HEAD` - `pnpm exec vitest run packages/adapter-utils/src/server-utils.test.ts packages/mcp-server/src/tools.test.ts packages/shared/src/issue-thread-interactions.test.ts ui/src/lib/issue-thread-interactions.test.ts ui/src/lib/issue-chat-messages.test.ts ui/src/components/IssueThreadInteractionCard.test.tsx ui/src/components/IssueChatThread.test.tsx server/src/__tests__/issue-thread-interaction-routes.test.ts server/src/__tests__/issue-thread-interactions-service.test.ts server/src/services/issue-thread-interactions.test.ts` -> 9 files / 79 tests passed - `pnpm -r typecheck` -> passed, including `packages/db` migration numbering check ## Risks - Medium: this adds a new issue-thread interaction model across db/shared/server/ui/plugin surfaces. - Migration risk is reduced by placing this branch after current master migrations (`0063`, `0064`) and making the idempotency column add idempotent for users who applied the old branch numbering. - UI interaction behavior is covered by component tests, but this PR does not include browser screenshots. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5-class coding agent runtime. Exact model ID and context window are not exposed in this Paperclip run; tool use and local shell/code execution were enabled. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-21 20:15:11 -05:00
IssueThreadInteraction,
CreateIssueThreadInteraction,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
IssueDocument,
2026-03-13 16:22:34 -05:00
Agent,
Goal,
} from "@paperclipai/shared";
import type {
EventFilter,
PluginContext,
PluginEntityRecord,
PluginEntityUpsert,
PluginJobContext,
PluginLauncherRegistration,
PluginEvent,
ScopeKey,
ToolResult,
ToolRunContext,
PluginWorkspace,
AgentSession,
AgentSessionEvent,
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
PluginLocalFolderEntry,
PluginLocalFolderStatus,
2026-03-13 16:22:34 -05:00
} from "./types.js";
Add sandbox environment support (#4415) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The environment/runtime layer decides where agent work executes and how the control plane reaches those runtimes. > - Today Paperclip can run locally and over SSH, but sandboxed execution needs a first-class environment model instead of one-off adapter behavior. > - We also want sandbox providers to be pluggable so the core does not hardcode every provider implementation. > - This branch adds the Sandbox environment path, the provider contract, and a deterministic fake provider plugin. > - That required synchronized changes across shared contracts, plugin SDK surfaces, server runtime orchestration, and the UI environment/workspace flows. > - The result is that sandbox execution becomes a core control-plane capability while keeping provider implementations extensible and testable. ## What Changed - Added sandbox runtime support to the environment execution path, including runtime URL discovery, sandbox execution targeting, orchestration, and heartbeat integration. - Added plugin-provider support for sandbox environments so providers can be supplied via plugins instead of hardcoded server logic. - Added the fake sandbox provider plugin with deterministic behavior suitable for local and automated testing. - Updated shared types, validators, plugin protocol definitions, and SDK helpers to carry sandbox provider and workspace-runtime contracts across package boundaries. - Updated server routes and services so companies can create sandbox environments, select them for work, and execute work through the sandbox runtime path. - Updated the UI environment and workspace surfaces to expose sandbox environment configuration and selection. - Added test coverage for sandbox runtime behavior, provider seams, environment route guards, orchestration, and the fake provider plugin. ## Verification - Ran locally before the final fixture-only scrub: - `pnpm -r typecheck` - `pnpm test:run` - `pnpm build` - Ran locally after the final scrub amend: - `pnpm vitest run server/src/__tests__/runtime-api.test.ts` - Reviewer spot checks: - create a sandbox environment backed by the fake provider plugin - run work through that environment - confirm sandbox provider execution does not inherit host secrets implicitly ## Risks - This touches shared contracts, plugin SDK plumbing, server runtime orchestration, and UI environment/workspace flows, so regressions would likely show up as cross-layer mismatches rather than isolated type errors. - Runtime URL discovery and sandbox callback selection are sensitive to host/bind configuration; if that logic is wrong, sandbox-backed callbacks may fail even when execution succeeds. - The fake provider plugin is intentionally deterministic and test-oriented; future providers may expose capability gaps that this branch does not yet cover. ## Model Used - OpenAI Codex coding agent on a GPT-5-class backend in the Paperclip/Codex harness. Exact backend model ID is not exposed in-session. Tool-assisted workflow with shell execution, file editing, git history inspection, and local test execution. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge
2026-04-24 12:15:53 -07:00
import type {
PluginEnvironmentValidateConfigParams,
PluginEnvironmentValidationResult,
PluginEnvironmentProbeParams,
PluginEnvironmentProbeResult,
PluginEnvironmentLease,
PluginEnvironmentAcquireLeaseParams,
PluginEnvironmentResumeLeaseParams,
PluginEnvironmentReleaseLeaseParams,
PluginEnvironmentDestroyLeaseParams,
PluginEnvironmentRealizeWorkspaceParams,
PluginEnvironmentRealizeWorkspaceResult,
PluginEnvironmentExecuteParams,
PluginEnvironmentExecuteResult,
} from "./protocol.js";
2026-03-13 16:22:34 -05:00
export interface TestHarnessOptions {
/** Plugin manifest used to seed capability checks and metadata. */
manifest: PaperclipPluginManifestV1;
/** Optional capability override. Defaults to `manifest.capabilities`. */
capabilities?: PluginCapability[];
/** Initial config returned by `ctx.config.get()`. */
config?: Record<string, unknown>;
}
export interface TestHarnessLogEntry {
level: "info" | "warn" | "error" | "debug";
message: string;
meta?: Record<string, unknown>;
}
export interface TestHarness {
/** Fully-typed in-memory plugin context passed to `plugin.setup(ctx)`. */
ctx: PluginContext;
/** Seed host entities for `ctx.companies/projects/issues/agents/goals` reads. */
seed(input: {
companies?: Company[];
projects?: Project[];
issues?: Issue[];
issueComments?: IssueComment[];
agents?: Agent[];
goals?: Goal[];
}): void;
setConfig(config: Record<string, unknown>): void;
/** Dispatch a host or plugin event to registered handlers. */
emit(eventType: PluginEventType | `plugin.${string}`, payload: unknown, base?: Partial<PluginEvent>): Promise<void>;
/** Execute a previously-registered scheduled job handler. */
runJob(jobKey: string, partial?: Partial<PluginJobContext>): Promise<void>;
/** Invoke a `ctx.data.register(...)` handler by key. */
getData<T = unknown>(key: string, params?: Record<string, unknown>): Promise<T>;
/** Invoke a `ctx.actions.register(...)` handler by key. */
performAction<T = unknown>(key: string, params?: Record<string, unknown>): Promise<T>;
/** Execute a registered tool handler via `ctx.tools.execute(...)`. */
executeTool<T = ToolResult>(name: string, params: unknown, runCtx?: Partial<ToolRunContext>): Promise<T>;
/** Read raw in-memory state for assertions. */
getState(input: ScopeKey): unknown;
/** Simulate a streaming event arriving for an active session. */
simulateSessionEvent(sessionId: string, event: Omit<AgentSessionEvent, "sessionId">): void;
logs: TestHarnessLogEntry[];
activity: Array<{ message: string; entityType?: string; entityId?: string; metadata?: Record<string, unknown> }>;
metrics: Array<{ name: string; value: number; tags?: Record<string, string> }>;
telemetry: Array<{ eventName: string; dimensions?: Record<string, string | number | boolean> }>;
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
dbQueries: Array<{ sql: string; params?: unknown[] }>;
dbExecutes: Array<{ sql: string; params?: unknown[] }>;
2026-03-13 16:22:34 -05:00
}
Add sandbox environment support (#4415) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The environment/runtime layer decides where agent work executes and how the control plane reaches those runtimes. > - Today Paperclip can run locally and over SSH, but sandboxed execution needs a first-class environment model instead of one-off adapter behavior. > - We also want sandbox providers to be pluggable so the core does not hardcode every provider implementation. > - This branch adds the Sandbox environment path, the provider contract, and a deterministic fake provider plugin. > - That required synchronized changes across shared contracts, plugin SDK surfaces, server runtime orchestration, and the UI environment/workspace flows. > - The result is that sandbox execution becomes a core control-plane capability while keeping provider implementations extensible and testable. ## What Changed - Added sandbox runtime support to the environment execution path, including runtime URL discovery, sandbox execution targeting, orchestration, and heartbeat integration. - Added plugin-provider support for sandbox environments so providers can be supplied via plugins instead of hardcoded server logic. - Added the fake sandbox provider plugin with deterministic behavior suitable for local and automated testing. - Updated shared types, validators, plugin protocol definitions, and SDK helpers to carry sandbox provider and workspace-runtime contracts across package boundaries. - Updated server routes and services so companies can create sandbox environments, select them for work, and execute work through the sandbox runtime path. - Updated the UI environment and workspace surfaces to expose sandbox environment configuration and selection. - Added test coverage for sandbox runtime behavior, provider seams, environment route guards, orchestration, and the fake provider plugin. ## Verification - Ran locally before the final fixture-only scrub: - `pnpm -r typecheck` - `pnpm test:run` - `pnpm build` - Ran locally after the final scrub amend: - `pnpm vitest run server/src/__tests__/runtime-api.test.ts` - Reviewer spot checks: - create a sandbox environment backed by the fake provider plugin - run work through that environment - confirm sandbox provider execution does not inherit host secrets implicitly ## Risks - This touches shared contracts, plugin SDK plumbing, server runtime orchestration, and UI environment/workspace flows, so regressions would likely show up as cross-layer mismatches rather than isolated type errors. - Runtime URL discovery and sandbox callback selection are sensitive to host/bind configuration; if that logic is wrong, sandbox-backed callbacks may fail even when execution succeeds. - The fake provider plugin is intentionally deterministic and test-oriented; future providers may expose capability gaps that this branch does not yet cover. ## Model Used - OpenAI Codex coding agent on a GPT-5-class backend in the Paperclip/Codex harness. Exact backend model ID is not exposed in-session. Tool-assisted workflow with shell execution, file editing, git history inspection, and local test execution. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge
2026-04-24 12:15:53 -07:00
// ---------------------------------------------------------------------------
// Environment test harness types
// ---------------------------------------------------------------------------
/** Recorded environment lifecycle event for assertion helpers. */
export interface EnvironmentEventRecord {
type:
| "validateConfig"
| "probe"
| "acquireLease"
| "resumeLease"
| "releaseLease"
| "destroyLease"
| "realizeWorkspace"
| "execute";
driverKey: string;
environmentId: string;
timestamp: string;
params: Record<string, unknown>;
result?: unknown;
error?: string;
}
/** Options for creating an environment-aware test harness. */
export interface EnvironmentTestHarnessOptions extends TestHarnessOptions {
/** Environment driver hooks provided by the plugin under test. */
environmentDriver: {
driverKey: string;
onValidateConfig?: (params: PluginEnvironmentValidateConfigParams) => Promise<PluginEnvironmentValidationResult>;
onProbe?: (params: PluginEnvironmentProbeParams) => Promise<PluginEnvironmentProbeResult>;
onAcquireLease?: (params: PluginEnvironmentAcquireLeaseParams) => Promise<PluginEnvironmentLease>;
onResumeLease?: (params: PluginEnvironmentResumeLeaseParams) => Promise<PluginEnvironmentLease>;
onReleaseLease?: (params: PluginEnvironmentReleaseLeaseParams) => Promise<void>;
onDestroyLease?: (params: PluginEnvironmentDestroyLeaseParams) => Promise<void>;
onRealizeWorkspace?: (params: PluginEnvironmentRealizeWorkspaceParams) => Promise<PluginEnvironmentRealizeWorkspaceResult>;
onExecute?: (params: PluginEnvironmentExecuteParams) => Promise<PluginEnvironmentExecuteResult>;
};
}
/** Extended test harness with environment driver simulation. */
export interface EnvironmentTestHarness extends TestHarness {
/** Recorded environment lifecycle events for assertion. */
environmentEvents: EnvironmentEventRecord[];
/** Invoke the environment driver's validateConfig hook. */
validateConfig(params: PluginEnvironmentValidateConfigParams): Promise<PluginEnvironmentValidationResult>;
/** Invoke the environment driver's probe hook. */
probe(params: PluginEnvironmentProbeParams): Promise<PluginEnvironmentProbeResult>;
/** Invoke the environment driver's acquireLease hook. */
acquireLease(params: PluginEnvironmentAcquireLeaseParams): Promise<PluginEnvironmentLease>;
/** Invoke the environment driver's resumeLease hook. */
resumeLease(params: PluginEnvironmentResumeLeaseParams): Promise<PluginEnvironmentLease>;
/** Invoke the environment driver's releaseLease hook. */
releaseLease(params: PluginEnvironmentReleaseLeaseParams): Promise<void>;
/** Invoke the environment driver's destroyLease hook. */
destroyLease(params: PluginEnvironmentDestroyLeaseParams): Promise<void>;
/** Invoke the environment driver's realizeWorkspace hook. */
realizeWorkspace(params: PluginEnvironmentRealizeWorkspaceParams): Promise<PluginEnvironmentRealizeWorkspaceResult>;
/** Invoke the environment driver's execute hook. */
execute(params: PluginEnvironmentExecuteParams): Promise<PluginEnvironmentExecuteResult>;
}
// ---------------------------------------------------------------------------
// Environment event assertion helpers
// ---------------------------------------------------------------------------
/** Filter environment events by type. */
export function filterEnvironmentEvents(
events: EnvironmentEventRecord[],
type: EnvironmentEventRecord["type"],
): EnvironmentEventRecord[] {
return events.filter((e) => e.type === type);
}
/** Assert that environment events occurred in the expected order. */
export function assertEnvironmentEventOrder(
events: EnvironmentEventRecord[],
expectedOrder: EnvironmentEventRecord["type"][],
): void {
const actual = events.map((e) => e.type);
const matched: EnvironmentEventRecord["type"][] = [];
let cursor = 0;
for (const eventType of actual) {
if (cursor < expectedOrder.length && eventType === expectedOrder[cursor]) {
matched.push(eventType);
cursor++;
}
}
if (matched.length !== expectedOrder.length) {
throw new Error(
`Environment event order mismatch.\nExpected: ${JSON.stringify(expectedOrder)}\nActual: ${JSON.stringify(actual)}`,
);
}
}
/** Assert that a full lease lifecycle (acquire → release) occurred for an environment. */
export function assertLeaseLifecycle(
events: EnvironmentEventRecord[],
environmentId: string,
): { acquire: EnvironmentEventRecord; release: EnvironmentEventRecord } {
const acquire = events.find((e) => e.type === "acquireLease" && e.environmentId === environmentId);
const release = events.find((e) => (e.type === "releaseLease" || e.type === "destroyLease") && e.environmentId === environmentId);
if (!acquire) throw new Error(`No acquireLease event found for environment ${environmentId}`);
if (!release) throw new Error(`No releaseLease/destroyLease event found for environment ${environmentId}`);
if (acquire.timestamp > release.timestamp) {
throw new Error(`acquireLease occurred after release for environment ${environmentId}`);
}
return { acquire, release };
}
/** Assert that workspace realization occurred between lease acquire and release. */
export function assertWorkspaceRealizationLifecycle(
events: EnvironmentEventRecord[],
environmentId: string,
): EnvironmentEventRecord {
const lifecycle = assertLeaseLifecycle(events, environmentId);
const realize = events.find(
(e) => e.type === "realizeWorkspace" && e.environmentId === environmentId,
);
if (!realize) throw new Error(`No realizeWorkspace event found for environment ${environmentId}`);
if (realize.timestamp < lifecycle.acquire.timestamp) {
throw new Error(`realizeWorkspace occurred before acquireLease for environment ${environmentId}`);
}
if (realize.timestamp > lifecycle.release.timestamp) {
throw new Error(`realizeWorkspace occurred after release for environment ${environmentId}`);
}
return realize;
}
/** Assert that an execute call occurred within the lease lifecycle. */
export function assertExecutionLifecycle(
events: EnvironmentEventRecord[],
environmentId: string,
): EnvironmentEventRecord[] {
const lifecycle = assertLeaseLifecycle(events, environmentId);
const execEvents = events.filter(
(e) => e.type === "execute" && e.environmentId === environmentId,
);
if (execEvents.length === 0) {
throw new Error(`No execute events found for environment ${environmentId}`);
}
for (const exec of execEvents) {
if (exec.timestamp < lifecycle.acquire.timestamp || exec.timestamp > lifecycle.release.timestamp) {
throw new Error(`Execute event occurred outside lease lifecycle for environment ${environmentId}`);
}
}
return execEvents;
}
/** Assert that an event recorded an error. */
export function assertEnvironmentError(
events: EnvironmentEventRecord[],
type: EnvironmentEventRecord["type"],
environmentId?: string,
): EnvironmentEventRecord {
const match = events.find(
(e) => e.type === type && e.error != null && (!environmentId || e.environmentId === environmentId),
);
if (!match) {
throw new Error(`No error event of type '${type}'${environmentId ? ` for environment ${environmentId}` : ""}`);
}
return match;
}
// ---------------------------------------------------------------------------
// Fake environment plugin driver
// ---------------------------------------------------------------------------
/** Options for creating a fake environment driver for contract testing. */
export interface FakeEnvironmentDriverOptions {
driverKey?: string;
/** Simulated acquire delay in ms. */
acquireDelayMs?: number;
/** If true, probe will return `ok: false`. */
probeFailure?: boolean;
/** If true, acquireLease will throw. */
acquireFailure?: string;
/** If true, execute will return a non-zero exit code. */
executeFailure?: boolean;
/** Custom metadata returned on lease acquire. */
leaseMetadata?: Record<string, unknown>;
}
/**
* Create a fake environment driver suitable for contract testing.
*
* This returns a driver hooks object compatible with `EnvironmentTestHarnessOptions.environmentDriver`.
* It simulates the full environment lifecycle with configurable failure injection.
*/
export function createFakeEnvironmentDriver(options: FakeEnvironmentDriverOptions = {}): EnvironmentTestHarnessOptions["environmentDriver"] {
const driverKey = options.driverKey ?? "fake";
const leases = new Map<string, { providerLeaseId: string; metadata: Record<string, unknown> }>();
let leaseCounter = 0;
return {
driverKey,
async onValidateConfig(params) {
if (!params.config || typeof params.config !== "object") {
return { ok: false, errors: ["Config must be an object"] };
}
return { ok: true, normalizedConfig: params.config };
},
async onProbe(_params) {
if (options.probeFailure) {
return { ok: false, summary: "Simulated probe failure", diagnostics: [{ severity: "error", message: "Probe failed" }] };
}
return { ok: true, summary: "Fake environment is healthy" };
},
async onAcquireLease(params) {
if (options.acquireFailure) {
throw new Error(options.acquireFailure);
}
if (options.acquireDelayMs) {
await new Promise((resolve) => setTimeout(resolve, options.acquireDelayMs));
}
const providerLeaseId = `fake-lease-${++leaseCounter}`;
const metadata = { ...options.leaseMetadata, acquiredAt: new Date().toISOString(), runId: params.runId };
leases.set(providerLeaseId, { providerLeaseId, metadata });
return { providerLeaseId, metadata };
},
async onResumeLease(params) {
const existing = leases.get(params.providerLeaseId);
if (!existing) {
throw new Error(`Lease ${params.providerLeaseId} not found — cannot resume`);
}
return { providerLeaseId: existing.providerLeaseId, metadata: { ...existing.metadata, resumed: true } };
},
async onReleaseLease(params) {
if (params.providerLeaseId) {
leases.delete(params.providerLeaseId);
}
},
async onDestroyLease(params) {
if (params.providerLeaseId) {
leases.delete(params.providerLeaseId);
}
},
async onRealizeWorkspace(params) {
return {
cwd: params.workspace.localPath ?? params.workspace.remotePath ?? "/tmp/fake-workspace",
metadata: { realized: true },
};
},
async onExecute(params) {
if (options.executeFailure) {
return { exitCode: 1, timedOut: false, stdout: "", stderr: "Simulated execution failure" };
}
return {
exitCode: 0,
timedOut: false,
stdout: `Executed: ${params.command} ${(params.args ?? []).join(" ")}`.trim(),
stderr: "",
};
},
};
}
2026-03-13 16:22:34 -05:00
type EventRegistration = {
name: PluginEventType | `plugin.${string}`;
filter?: EventFilter;
fn: (event: PluginEvent) => Promise<void>;
};
function normalizeScope(input: ScopeKey): Required<Pick<ScopeKey, "scopeKind" | "stateKey">> & Pick<ScopeKey, "scopeId" | "namespace"> {
return {
scopeKind: input.scopeKind,
scopeId: input.scopeId,
namespace: input.namespace ?? "default",
stateKey: input.stateKey,
};
}
function stateMapKey(input: ScopeKey): string {
const normalized = normalizeScope(input);
return `${normalized.scopeKind}|${normalized.scopeId ?? ""}|${normalized.namespace}|${normalized.stateKey}`;
}
function allowsEvent(filter: EventFilter | undefined, event: PluginEvent): boolean {
if (!filter) return true;
if (filter.companyId && filter.companyId !== String((event.payload as Record<string, unknown> | undefined)?.companyId ?? "")) return false;
if (filter.projectId && filter.projectId !== String((event.payload as Record<string, unknown> | undefined)?.projectId ?? "")) return false;
if (filter.agentId && filter.agentId !== String((event.payload as Record<string, unknown> | undefined)?.agentId ?? "")) return false;
return true;
}
function requireCapability(manifest: PaperclipPluginManifestV1, allowed: Set<PluginCapability>, capability: PluginCapability) {
if (allowed.has(capability)) return;
throw new Error(`Plugin '${manifest.id}' is missing required capability '${capability}' in test harness`);
}
function requireCompanyId(companyId?: string): string {
if (!companyId) throw new Error("companyId is required for this operation");
return companyId;
}
function isInCompany<T extends { companyId: string | null | undefined }>(
record: T | null | undefined,
companyId: string,
): record is T {
return Boolean(record && record.companyId === companyId);
}
/**
* Create an in-memory host harness for plugin worker tests.
*
* The harness enforces declared capabilities and simulates host APIs, so tests
* can validate plugin behavior without spinning up the Paperclip server runtime.
*/
export function createTestHarness(options: TestHarnessOptions): TestHarness {
const manifest = options.manifest;
const capabilitySet = new Set(options.capabilities ?? manifest.capabilities);
let currentConfig = { ...(options.config ?? {}) };
const logs: TestHarnessLogEntry[] = [];
const activity: TestHarness["activity"] = [];
const metrics: TestHarness["metrics"] = [];
const telemetry: TestHarness["telemetry"] = [];
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
const dbQueries: TestHarness["dbQueries"] = [];
const dbExecutes: TestHarness["dbExecutes"] = [];
2026-03-13 16:22:34 -05:00
const state = new Map<string, unknown>();
const entities = new Map<string, PluginEntityRecord>();
const entityExternalIndex = new Map<string, string>();
const companies = new Map<string, Company>();
const projects = new Map<string, Project>();
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
const routines = new Map<string, Routine>();
const routineRuns = new Map<string, RoutineRun>();
2026-03-13 16:22:34 -05:00
const issues = new Map<string, Issue>();
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
const blockedByIssueIds = new Map<string, string[]>();
2026-03-13 16:22:34 -05:00
const issueComments = new Map<string, IssueComment[]>();
[codex] Add structured issue-thread interactions (#4244) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - Operators supervise that work through issues, comments, approvals, and the board UI. > - Some agent proposals need structured board/user decisions, not hidden markdown conventions or heavyweight governed approvals. > - Issue-thread interactions already provide a natural thread-native surface for proposed tasks and questions. > - This pull request extends that surface with request confirmations, richer interaction cards, and agent/plugin/MCP helpers. > - The benefit is that plan approvals and yes/no decisions become explicit, auditable, and resumable without losing the single-issue workflow. ## What Changed - Added persisted issue-thread interactions for suggested tasks, structured questions, and request confirmations. - Added board UI cards for interaction review, selection, question answers, and accept/reject confirmation flows. - Added MCP and plugin SDK helpers for creating interaction cards from agents/plugins. - Updated agent wake instructions, onboarding assets, Paperclip skill docs, and public docs to prefer structured confirmations for issue-scoped decisions. - Rebased the branch onto `public-gh/master` and renumbered branch migrations to `0063` and `0064`; the idempotency migration uses `ADD COLUMN IF NOT EXISTS` for old branch users. ## Verification - `git diff --check public-gh/master..HEAD` - `pnpm exec vitest run packages/adapter-utils/src/server-utils.test.ts packages/mcp-server/src/tools.test.ts packages/shared/src/issue-thread-interactions.test.ts ui/src/lib/issue-thread-interactions.test.ts ui/src/lib/issue-chat-messages.test.ts ui/src/components/IssueThreadInteractionCard.test.tsx ui/src/components/IssueChatThread.test.tsx server/src/__tests__/issue-thread-interaction-routes.test.ts server/src/__tests__/issue-thread-interactions-service.test.ts server/src/services/issue-thread-interactions.test.ts` -> 9 files / 79 tests passed - `pnpm -r typecheck` -> passed, including `packages/db` migration numbering check ## Risks - Medium: this adds a new issue-thread interaction model across db/shared/server/ui/plugin surfaces. - Migration risk is reduced by placing this branch after current master migrations (`0063`, `0064`) and making the idempotency column add idempotent for users who applied the old branch numbering. - UI interaction behavior is covered by component tests, but this PR does not include browser screenshots. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5-class coding agent runtime. Exact model ID and context window are not exposed in this Paperclip run; tool use and local shell/code execution were enabled. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-21 20:15:11 -05:00
const issueInteractions = new Map<string, IssueThreadInteraction[]>();
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
const issueDocuments = new Map<string, IssueDocument>();
2026-03-13 16:22:34 -05:00
const agents = new Map<string, Agent>();
const goals = new Map<string, Goal>();
const projectWorkspaces = new Map<string, PluginWorkspace[]>();
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
const localFolderStatuses = new Map<string, PluginLocalFolderStatus>();
const localFolderFiles = new Map<string, string>();
2026-03-13 16:22:34 -05:00
const sessions = new Map<string, AgentSession>();
const sessionEventCallbacks = new Map<string, (event: AgentSessionEvent) => void>();
const events: EventRegistration[] = [];
const jobs = new Map<string, (job: PluginJobContext) => Promise<void>>();
const launchers = new Map<string, PluginLauncherRegistration>();
const dataHandlers = new Map<string, (params: Record<string, unknown>) => Promise<unknown>>();
const actionHandlers = new Map<string, (params: Record<string, unknown>) => Promise<unknown>>();
const toolHandlers = new Map<string, (params: unknown, runCtx: ToolRunContext) => Promise<ToolResult>>();
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
function localFolderKey(companyId: string, folderKey: string): string {
return `${companyId}:${folderKey}`;
}
function localFolderFileKey(companyId: string, folderKey: string, relativePath: string): string {
return `${localFolderKey(companyId, folderKey)}:${relativePath}`;
}
function normalizeLocalFolderRelativePath(relativePath: string): string {
const parts: string[] = [];
for (const segment of relativePath.split(/[\\/]+/)) {
if (!segment || segment === ".") continue;
if (segment === "..") throw new Error("Local folder path traversal is not allowed");
parts.push(segment);
}
return parts.join("/");
}
function notConfiguredLocalFolderStatus(folderKey: string): PluginLocalFolderStatus {
return {
folderKey,
configured: false,
path: null,
realPath: null,
access: "readWrite",
readable: false,
writable: false,
requiredDirectories: [],
requiredFiles: [],
missingDirectories: [],
missingFiles: [],
healthy: false,
problems: [{ code: "not_configured", message: "No local folder path is configured." }],
checkedAt: new Date().toISOString(),
};
}
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
function issueRelationSummary(issueId: string) {
const issue = issues.get(issueId);
if (!issue) throw new Error(`Issue not found: ${issueId}`);
const summarize = (candidateId: string) => {
const related = issues.get(candidateId);
if (!related || related.companyId !== issue.companyId) return null;
return {
id: related.id,
identifier: related.identifier,
title: related.title,
status: related.status,
priority: related.priority,
assigneeAgentId: related.assigneeAgentId,
assigneeUserId: related.assigneeUserId,
};
};
const blockedBy = (blockedByIssueIds.get(issueId) ?? [])
.map(summarize)
.filter((value): value is NonNullable<typeof value> => value !== null);
const blocks = [...blockedByIssueIds.entries()]
.filter(([, blockers]) => blockers.includes(issueId))
.map(([blockedIssueId]) => summarize(blockedIssueId))
.filter((value): value is NonNullable<typeof value> => value !== null);
return { blockedBy, blocks };
}
const defaultPluginOriginKind: PluginIssueOriginKind = `plugin:${manifest.id}`;
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
function managedAgentDeclaration(agentKey: string) {
const declaration = manifest.agents?.find((agent) => agent.agentKey === agentKey);
if (!declaration) throw new Error(`Managed agent declaration not found: ${agentKey}`);
return declaration;
}
function isManagedAgent(agent: Agent, agentKey: string) {
const marker = agent.metadata?.paperclipManagedResource;
return Boolean(
marker
&& typeof marker === "object"
&& !Array.isArray(marker)
&& (marker as Record<string, unknown>).pluginKey === manifest.id
&& (marker as Record<string, unknown>).resourceKind === "agent"
&& (marker as Record<string, unknown>).resourceKey === agentKey,
);
}
function managedAgentMetadata(agentKey: string, existing?: Record<string, unknown> | null) {
return {
...(existing ?? {}),
paperclipManagedResource: {
pluginKey: manifest.id,
resourceKind: "agent",
resourceKey: agentKey,
},
};
}
function managedResolution(
agentKey: string,
companyId: string,
agent: Agent | null,
status: PluginManagedAgentResolution["status"],
): PluginManagedAgentResolution {
return {
pluginKey: manifest.id,
resourceKind: "agent",
resourceKey: agentKey,
companyId,
agentId: agent?.id ?? null,
agent,
status,
approvalId: null,
};
}
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
function normalizePluginOriginKind(originKind: unknown = defaultPluginOriginKind): PluginIssueOriginKind {
if (originKind == null || originKind === "") return defaultPluginOriginKind;
if (typeof originKind !== "string") throw new Error("Plugin issue originKind must be a string");
if (originKind === defaultPluginOriginKind || originKind.startsWith(`${defaultPluginOriginKind}:`)) {
return originKind as PluginIssueOriginKind;
}
throw new Error(`Plugin may only use originKind values under ${defaultPluginOriginKind}`);
}
2026-03-13 16:22:34 -05:00
const ctx: PluginContext = {
manifest,
config: {
async get() {
return { ...currentConfig };
},
},
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
localFolders: {
declarations() {
return manifest.localFolders ?? [];
},
async configure(input) {
requireCapability(manifest, capabilitySet, "local.folders");
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
const status = {
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
folderKey: input.folderKey,
configured: true,
path: input.path,
realPath: input.path,
access: input.access ?? "readWrite",
readable: true,
writable: input.access === "read" ? false : true,
requiredDirectories: input.requiredDirectories ?? [],
requiredFiles: input.requiredFiles ?? [],
missingDirectories: [],
missingFiles: [],
healthy: true,
problems: [],
checkedAt: new Date().toISOString(),
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
} satisfies PluginLocalFolderStatus;
localFolderStatuses.set(localFolderKey(input.companyId, input.folderKey), status);
return status;
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
},
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
async status(companyId, folderKey) {
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
requireCapability(manifest, capabilitySet, "local.folders");
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
return localFolderStatuses.get(localFolderKey(companyId, folderKey)) ?? notConfiguredLocalFolderStatus(folderKey);
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
},
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
async list(companyId, folderKey, options) {
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
requireCapability(manifest, capabilitySet, "local.folders");
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
const status = localFolderStatuses.get(localFolderKey(companyId, folderKey));
if (!status?.configured) throw new Error("Local folder is not configured");
const prefix = normalizeLocalFolderRelativePath(options?.relativePath ?? "");
const prefixWithSlash = prefix ? `${prefix}/` : "";
const entries = new Map<string, PluginLocalFolderEntry>();
for (const [key, contents] of localFolderFiles) {
const filePrefix = `${localFolderKey(companyId, folderKey)}:`;
if (!key.startsWith(filePrefix)) continue;
const filePath = key.slice(filePrefix.length);
if (prefix && filePath !== prefix && !filePath.startsWith(prefixWithSlash)) continue;
const remainder = prefix ? filePath.slice(prefixWithSlash.length) : filePath;
const [name] = remainder.split("/");
if (!name) continue;
const entryPath = prefix ? `${prefix}/${name}` : name;
const isNested = remainder.includes("/");
if (!options?.recursive && isNested) {
entries.set(entryPath, {
path: entryPath,
name,
kind: "directory",
size: null,
modifiedAt: null,
});
continue;
}
entries.set(filePath, {
path: filePath,
name: filePath.split("/").pop() ?? filePath,
kind: "file",
size: Buffer.byteLength(contents, "utf8"),
modifiedAt: null,
});
}
const maxEntries = options?.maxEntries && options.maxEntries > 0 ? options.maxEntries : entries.size;
const allEntries = [...entries.values()].sort((a, b) => a.path.localeCompare(b.path));
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
return {
folderKey,
relativePath: options?.relativePath ?? null,
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
entries: allEntries.slice(0, maxEntries),
truncated: allEntries.length > maxEntries,
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
};
},
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
async readText(companyId, folderKey, relativePath) {
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
requireCapability(manifest, capabilitySet, "local.folders");
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
const normalizedPath = normalizeLocalFolderRelativePath(relativePath);
const contents = localFolderFiles.get(localFolderFileKey(companyId, folderKey, normalizedPath));
if (contents === undefined) throw new Error(`Local folder file not found: ${relativePath}`);
return contents;
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
},
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
async writeTextAtomic(companyId, folderKey, relativePath, contents) {
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
requireCapability(manifest, capabilitySet, "local.folders");
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
const status = localFolderStatuses.get(localFolderKey(companyId, folderKey)) ?? {
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
folderKey,
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
configured: true,
path: `memory://${manifest.id}/${companyId}/${folderKey}`,
realPath: `memory://${manifest.id}/${companyId}/${folderKey}`,
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
access: "readWrite",
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
readable: true,
writable: true,
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
requiredDirectories: [],
requiredFiles: [],
missingDirectories: [],
missingFiles: [],
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
healthy: true,
problems: [],
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
checkedAt: new Date().toISOString(),
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
} satisfies PluginLocalFolderStatus;
if (status.access !== "readWrite" || !status.writable) {
throw new Error("Local folder is not configured for writes");
}
localFolderStatuses.set(localFolderKey(companyId, folderKey), status);
localFolderFiles.set(localFolderFileKey(companyId, folderKey, normalizeLocalFolderRelativePath(relativePath)), contents);
return status;
},
async deleteFile(companyId, folderKey, relativePath) {
requireCapability(manifest, capabilitySet, "local.folders");
const status = localFolderStatuses.get(localFolderKey(companyId, folderKey)) ?? notConfiguredLocalFolderStatus(folderKey);
if (status.configured && (status.access !== "readWrite" || !status.writable)) {
throw new Error("Local folder is not configured for writes");
}
localFolderFiles.delete(localFolderFileKey(companyId, folderKey, normalizeLocalFolderRelativePath(relativePath)));
return status;
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
},
},
2026-03-13 16:22:34 -05:00
events: {
on(name: PluginEventType | `plugin.${string}`, filterOrFn: EventFilter | ((event: PluginEvent) => Promise<void>), maybeFn?: (event: PluginEvent) => Promise<void>): () => void {
requireCapability(manifest, capabilitySet, "events.subscribe");
let registration: EventRegistration;
if (typeof filterOrFn === "function") {
registration = { name, fn: filterOrFn };
} else {
if (!maybeFn) throw new Error("event handler is required");
registration = { name, filter: filterOrFn, fn: maybeFn };
}
events.push(registration);
return () => {
const idx = events.indexOf(registration);
if (idx !== -1) events.splice(idx, 1);
};
},
async emit(name, companyId, payload) {
requireCapability(manifest, capabilitySet, "events.emit");
await harness.emit(`plugin.${manifest.id}.${name}`, payload, { companyId });
},
},
jobs: {
register(key, fn) {
requireCapability(manifest, capabilitySet, "jobs.schedule");
jobs.set(key, fn);
},
},
launchers: {
register(launcher) {
launchers.set(launcher.id, launcher);
},
},
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
db: {
namespace: manifest.database ? `test_${manifest.id.replace(/[^a-z0-9_]+/g, "_")}` : "",
async query(sql, params) {
requireCapability(manifest, capabilitySet, "database.namespace.read");
dbQueries.push({ sql, params });
return [];
},
async execute(sql, params) {
requireCapability(manifest, capabilitySet, "database.namespace.write");
dbExecutes.push({ sql, params });
return { rowCount: 0 };
},
},
2026-03-13 16:22:34 -05:00
http: {
async fetch(url, init) {
requireCapability(manifest, capabilitySet, "http.outbound");
return fetch(url, init);
},
},
secrets: {
async resolve(secretRef) {
requireCapability(manifest, capabilitySet, "secrets.read-ref");
return `resolved:${secretRef}`;
},
},
activity: {
async log(entry) {
requireCapability(manifest, capabilitySet, "activity.log.write");
activity.push(entry);
},
},
state: {
async get(input) {
requireCapability(manifest, capabilitySet, "plugin.state.read");
return state.has(stateMapKey(input)) ? state.get(stateMapKey(input)) : null;
},
async set(input, value) {
requireCapability(manifest, capabilitySet, "plugin.state.write");
state.set(stateMapKey(input), value);
},
async delete(input) {
requireCapability(manifest, capabilitySet, "plugin.state.write");
state.delete(stateMapKey(input));
},
},
entities: {
async upsert(input: PluginEntityUpsert) {
const externalKey = input.externalId
? `${input.entityType}|${input.scopeKind}|${input.scopeId ?? ""}|${input.externalId}`
: null;
const existingId = externalKey ? entityExternalIndex.get(externalKey) : undefined;
const existing = existingId ? entities.get(existingId) : undefined;
const now = new Date().toISOString();
const previousExternalKey = existing?.externalId
? `${existing.entityType}|${existing.scopeKind}|${existing.scopeId ?? ""}|${existing.externalId}`
: null;
const record: PluginEntityRecord = existing
? {
...existing,
entityType: input.entityType,
scopeKind: input.scopeKind,
scopeId: input.scopeId ?? null,
externalId: input.externalId ?? null,
title: input.title ?? null,
status: input.status ?? null,
data: input.data,
updatedAt: now,
}
: {
id: randomUUID(),
entityType: input.entityType,
scopeKind: input.scopeKind,
scopeId: input.scopeId ?? null,
externalId: input.externalId ?? null,
title: input.title ?? null,
status: input.status ?? null,
data: input.data,
createdAt: now,
updatedAt: now,
};
entities.set(record.id, record);
if (previousExternalKey && previousExternalKey !== externalKey) {
entityExternalIndex.delete(previousExternalKey);
}
if (externalKey) entityExternalIndex.set(externalKey, record.id);
return record;
},
async list(query) {
let out = [...entities.values()];
if (query.entityType) out = out.filter((r) => r.entityType === query.entityType);
if (query.scopeKind) out = out.filter((r) => r.scopeKind === query.scopeKind);
if (query.scopeId) out = out.filter((r) => r.scopeId === query.scopeId);
if (query.externalId) out = out.filter((r) => r.externalId === query.externalId);
if (query.offset) out = out.slice(query.offset);
if (query.limit) out = out.slice(0, query.limit);
return out;
},
},
projects: {
async list(input) {
requireCapability(manifest, capabilitySet, "projects.read");
const companyId = requireCompanyId(input?.companyId);
let out = [...projects.values()];
out = out.filter((project) => project.companyId === companyId);
if (input?.offset) out = out.slice(input.offset);
if (input?.limit) out = out.slice(0, input.limit);
return out;
},
async get(projectId, companyId) {
requireCapability(manifest, capabilitySet, "projects.read");
const project = projects.get(projectId);
return isInCompany(project, companyId) ? project : null;
},
async listWorkspaces(projectId, companyId) {
requireCapability(manifest, capabilitySet, "project.workspaces.read");
if (!isInCompany(projects.get(projectId), companyId)) return [];
return projectWorkspaces.get(projectId) ?? [];
},
async getPrimaryWorkspace(projectId, companyId) {
requireCapability(manifest, capabilitySet, "project.workspaces.read");
if (!isInCompany(projects.get(projectId), companyId)) return null;
const workspaces = projectWorkspaces.get(projectId) ?? [];
return workspaces.find((workspace) => workspace.isPrimary) ?? null;
},
async getWorkspaceForIssue(issueId, companyId) {
requireCapability(manifest, capabilitySet, "project.workspaces.read");
const issue = issues.get(issueId);
if (!isInCompany(issue, companyId)) return null;
const projectId = (issue as unknown as Record<string, unknown>)?.projectId as string | undefined;
if (!projectId) return null;
if (!isInCompany(projects.get(projectId), companyId)) return null;
const workspaces = projectWorkspaces.get(projectId) ?? [];
return workspaces.find((workspace) => workspace.isPrimary) ?? null;
},
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
managed: {
async get(projectKey, companyId) {
requireCapability(manifest, capabilitySet, "projects.managed");
const declaration = manifest.projects?.find((project) => project.projectKey === projectKey);
if (!declaration) {
return {
pluginKey: manifest.id,
resourceKind: "project",
resourceKey: projectKey,
companyId,
projectId: null,
project: null,
status: "missing",
};
}
const externalId = `${manifest.id}:project:${projectKey}`;
const existingEntity = [...entities.values()].find((entity) =>
entity.entityType === "managed_resource"
&& entity.scopeKind === "company"
&& entity.scopeId === companyId
&& entity.externalId === externalId
);
const existingProject = existingEntity ? projects.get(String(existingEntity.data?.projectId ?? "")) : null;
if (existingProject && isInCompany(existingProject, companyId)) {
return {
pluginKey: manifest.id,
resourceKind: "project",
resourceKey: projectKey,
companyId,
projectId: existingProject.id,
project: existingProject,
status: "resolved",
};
}
const now = new Date();
const project = {
id: `project-${projects.size + 1}`,
companyId,
urlKey: declaration.projectKey,
goalId: null,
goalIds: [],
goals: [],
name: declaration.displayName,
description: declaration.description ?? null,
status: declaration.status ?? "in_progress",
leadAgentId: null,
targetDate: null,
color: declaration.color ?? null,
env: null,
pauseReason: null,
pausedAt: null,
executionWorkspacePolicy: null,
codebase: {
workspaceId: null,
repoUrl: null,
repoRef: null,
defaultRef: null,
repoName: null,
localFolder: null,
managedFolder: `/tmp/${declaration.projectKey}`,
effectiveLocalFolder: `/tmp/${declaration.projectKey}`,
origin: "managed_checkout",
},
workspaces: [],
primaryWorkspace: null,
managedByPlugin: {
id: `managed-${projects.size + 1}`,
pluginId: manifest.id,
pluginKey: manifest.id,
pluginDisplayName: manifest.displayName,
resourceKind: "project",
resourceKey: projectKey,
defaultsJson: { displayName: declaration.displayName, settings: declaration.settings ?? {} },
createdAt: now,
updatedAt: now,
},
archivedAt: null,
createdAt: now,
updatedAt: now,
} as Project;
projects.set(project.id, project);
const externalKey = `managed_resource|company|${companyId}|${externalId}`;
const nowIso = now.toISOString();
const record: PluginEntityRecord = {
id: randomUUID(),
entityType: "managed_resource",
scopeKind: "company",
scopeId: companyId,
externalId,
title: declaration.displayName,
status: null,
data: { resourceKind: "project", resourceKey: projectKey, projectId: project.id },
createdAt: nowIso,
updatedAt: nowIso,
};
entities.set(record.id, record);
entityExternalIndex.set(externalKey, record.id);
return {
pluginKey: manifest.id,
resourceKind: "project",
resourceKey: projectKey,
companyId,
projectId: project.id,
project,
status: "created",
};
},
async reconcile(projectKey, companyId) {
return this.get(projectKey, companyId);
},
async reset(projectKey, companyId) {
const resolved = await this.get(projectKey, companyId);
return { ...resolved, status: resolved.project ? "reset" : resolved.status };
},
},
},
routines: {
managed: {
async get(routineKey, companyId) {
requireCapability(manifest, capabilitySet, "routines.managed");
const declaration = manifest.routines?.find((routine) => routine.routineKey === routineKey);
if (!declaration) {
return {
pluginKey: manifest.id,
resourceKind: "routine",
resourceKey: routineKey,
companyId,
routineId: null,
routine: null,
status: "missing",
missingRefs: [],
} satisfies PluginManagedRoutineResolution;
}
const externalId = `${manifest.id}:routine:${routineKey}`;
const existingEntity = [...entities.values()].find((entity) =>
entity.entityType === "managed_resource"
&& entity.scopeKind === "company"
&& entity.scopeId === companyId
&& entity.externalId === externalId
);
const existingRoutine = existingEntity ? routines.get(String(existingEntity.data?.routineId ?? "")) : null;
if (existingRoutine && isInCompany(existingRoutine, companyId)) {
return {
pluginKey: manifest.id,
resourceKind: "routine",
resourceKey: routineKey,
companyId,
routineId: existingRoutine.id,
routine: existingRoutine,
status: "resolved",
missingRefs: [],
} satisfies PluginManagedRoutineResolution;
}
return {
pluginKey: manifest.id,
resourceKind: "routine",
resourceKey: routineKey,
companyId,
routineId: null,
routine: null,
status: "missing",
missingRefs: [],
} satisfies PluginManagedRoutineResolution;
},
async reconcile(routineKey, companyId, overrides) {
const existing = await this.get(routineKey, companyId);
if (existing.routine) return existing;
const declaration = manifest.routines?.find((routine) => routine.routineKey === routineKey);
if (!declaration) return existing;
const now = new Date();
const agentRef = declaration.assigneeRef;
const projectRef = declaration.projectRef;
const assigneeAgentId = overrides?.assigneeAgentId
?? (agentRef?.resourceKind === "agent"
? [...agents.values()].find((agent) => isInCompany(agent, companyId) && isManagedAgent(agent, agentRef.resourceKey))?.id
: null)
?? null;
const projectId = overrides?.projectId
?? (projectRef?.resourceKind === "project"
? [...projects.values()].find((project) => (
isInCompany(project, companyId)
&& project.managedByPlugin?.pluginKey === manifest.id
&& project.managedByPlugin?.resourceKey === projectRef.resourceKey
))?.id
: null)
?? null;
const missingRefs: NonNullable<PluginManagedRoutineResolution["missingRefs"]> = [];
if (agentRef && !assigneeAgentId) missingRefs.push({ ...agentRef, pluginKey: manifest.id });
if (projectRef && !projectId) missingRefs.push({ ...projectRef, pluginKey: manifest.id });
if (missingRefs.length > 0) {
return {
pluginKey: manifest.id,
resourceKind: "routine",
resourceKey: routineKey,
companyId,
routineId: null,
routine: null,
status: "missing_refs",
missingRefs,
} satisfies PluginManagedRoutineResolution;
}
const routine = {
id: `routine-${routines.size + 1}`,
companyId,
projectId,
goalId: declaration.goalId ?? null,
parentIssueId: null,
title: declaration.title,
description: declaration.description ?? null,
assigneeAgentId,
priority: declaration.priority ?? "medium",
status: declaration.status ?? (assigneeAgentId ? "active" : "paused"),
concurrencyPolicy: declaration.concurrencyPolicy ?? "coalesce_if_active",
catchUpPolicy: declaration.catchUpPolicy ?? "skip_missed",
variables: declaration.variables ?? [],
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
latestRevisionId: null,
latestRevisionNumber: 1,
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
createdByAgentId: null,
createdByUserId: null,
updatedByAgentId: null,
updatedByUserId: null,
lastTriggeredAt: null,
lastEnqueuedAt: null,
createdAt: now,
updatedAt: now,
managedByPlugin: {
id: `managed-routine-${routines.size + 1}`,
pluginId: manifest.id,
pluginKey: manifest.id,
pluginDisplayName: manifest.displayName,
resourceKind: "routine",
resourceKey: routineKey,
defaultsJson: { title: declaration.title, issueTemplate: declaration.issueTemplate ?? null },
createdAt: now,
updatedAt: now,
},
} as Routine;
routines.set(routine.id, routine);
const nowIso = now.toISOString();
const record: PluginEntityRecord = {
id: randomUUID(),
entityType: "managed_resource",
scopeKind: "company",
scopeId: companyId,
externalId: `${manifest.id}:routine:${routineKey}`,
title: declaration.title,
status: null,
data: { resourceKind: "routine", resourceKey: routineKey, routineId: routine.id },
createdAt: nowIso,
updatedAt: nowIso,
};
entities.set(record.id, record);
return {
pluginKey: manifest.id,
resourceKind: "routine",
resourceKey: routineKey,
companyId,
routineId: routine.id,
routine,
status: "created",
missingRefs: [],
} satisfies PluginManagedRoutineResolution;
},
async reset(routineKey, companyId, overrides) {
const resolved = await this.reconcile(routineKey, companyId, overrides);
return { ...resolved, status: resolved.routine ? "reset" : resolved.status } satisfies PluginManagedRoutineResolution;
},
async update(routineKey, companyId, patch) {
const resolved = await this.get(routineKey, companyId);
if (!resolved.routine) throw new Error(`Managed routine not found: ${routineKey}`);
const next = {
...resolved.routine,
...(patch.status !== undefined ? { status: patch.status } : {}),
updatedAt: new Date(),
};
routines.set(next.id, next);
return next;
},
async run(routineKey, companyId) {
const resolved = await this.get(routineKey, companyId);
if (!resolved.routine) throw new Error(`Managed routine not found: ${routineKey}`);
const now = new Date();
const run = {
id: `routine-run-${routineRuns.size + 1}`,
companyId,
routineId: resolved.routine.id,
triggerId: null,
source: "manual",
status: "queued",
triggeredAt: now,
idempotencyKey: null,
triggerPayload: null,
dispatchFingerprint: null,
linkedIssueId: null,
coalescedIntoRunId: null,
failureReason: null,
completedAt: null,
createdAt: now,
updatedAt: now,
} satisfies RoutineRun;
routineRuns.set(run.id, run);
routines.set(resolved.routine.id, {
...resolved.routine,
lastTriggeredAt: now,
lastEnqueuedAt: now,
updatedAt: now,
});
return run;
},
},
2026-03-13 16:22:34 -05:00
},
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
skills: {
managed: {
async get(skillKey, companyId) {
requireCapability(manifest, capabilitySet, "skills.managed");
const declaration = manifest.skills?.find((skill) => skill.skillKey === skillKey);
if (!declaration) {
return {
pluginKey: manifest.id,
resourceKind: "skill",
resourceKey: skillKey,
companyId,
skillId: null,
skill: null,
status: "missing",
defaultDrift: null,
} satisfies PluginManagedSkillResolution;
}
const externalId = `${manifest.id}:skill:${skillKey}`;
const existingEntity = [...entities.values()].find((entity) =>
entity.entityType === "managed_resource"
&& entity.scopeKind === "company"
&& entity.scopeId === companyId
&& entity.externalId === externalId
);
const existingSkill = existingEntity?.data?.skill as CompanySkill | undefined;
if (existingSkill && existingSkill.companyId === companyId) {
return {
pluginKey: manifest.id,
resourceKind: "skill",
resourceKey: skillKey,
companyId,
skillId: existingSkill.id,
skill: existingSkill,
status: "resolved",
defaultDrift: null,
} satisfies PluginManagedSkillResolution;
}
return {
pluginKey: manifest.id,
resourceKind: "skill",
resourceKey: skillKey,
companyId,
skillId: null,
skill: null,
status: "missing",
defaultDrift: null,
} satisfies PluginManagedSkillResolution;
},
async reconcile(skillKey, companyId) {
const existing = await this.get(skillKey, companyId);
if (existing.skill) return existing;
const declaration = manifest.skills?.find((skill) => skill.skillKey === skillKey);
if (!declaration) return existing;
const now = new Date();
const skill = {
id: randomUUID(),
companyId,
key: `plugin/${manifest.id.replace(/[^a-z0-9]+/g, "-").replace(/^-|-$/g, "")}/${skillKey}`,
slug: declaration.slug ?? skillKey,
name: declaration.displayName,
description: declaration.description ?? null,
markdown: declaration.markdown ?? `# ${declaration.displayName}\n`,
sourceType: "catalog",
sourceLocator: null,
sourceRef: null,
trustLevel: "markdown_only",
compatibility: "compatible",
fileInventory: [{ path: "SKILL.md", kind: "skill" }],
metadata: {
sourceKind: "catalog",
pluginManagedResource: {
pluginKey: manifest.id,
resourceKind: "skill",
resourceKey: skillKey,
},
},
createdAt: now,
updatedAt: now,
} satisfies CompanySkill;
const nowIso = now.toISOString();
const record: PluginEntityRecord = {
id: randomUUID(),
entityType: "managed_resource",
scopeKind: "company",
scopeId: companyId,
externalId: `${manifest.id}:skill:${skillKey}`,
title: declaration.displayName,
status: null,
data: { resourceKind: "skill", resourceKey: skillKey, skillId: skill.id, skill },
createdAt: nowIso,
updatedAt: nowIso,
};
entities.set(record.id, record);
return {
pluginKey: manifest.id,
resourceKind: "skill",
resourceKey: skillKey,
companyId,
skillId: skill.id,
skill,
status: "created",
defaultDrift: null,
} satisfies PluginManagedSkillResolution;
},
async reset(skillKey, companyId) {
requireCapability(manifest, capabilitySet, "skills.managed");
const existing = await this.get(skillKey, companyId);
const declaration = manifest.skills?.find((skill) => skill.skillKey === skillKey);
if (!declaration) return existing;
const now = new Date();
const skill = {
id: existing.skill?.id ?? randomUUID(),
companyId,
key: `plugin/${manifest.id.replace(/[^a-z0-9]+/g, "-").replace(/^-|-$/g, "")}/${skillKey}`,
slug: declaration.slug ?? skillKey,
name: declaration.displayName,
description: declaration.description ?? null,
markdown: declaration.markdown ?? `# ${declaration.displayName}\n`,
sourceType: "catalog",
sourceLocator: null,
sourceRef: null,
trustLevel: "markdown_only",
compatibility: "compatible",
fileInventory: [{ path: "SKILL.md", kind: "skill" }],
metadata: {
sourceKind: "catalog",
pluginManagedResource: {
pluginKey: manifest.id,
resourceKind: "skill",
resourceKey: skillKey,
},
},
createdAt: existing.skill?.createdAt ?? now,
updatedAt: now,
} satisfies CompanySkill;
const nowIso = now.toISOString();
const existingEntity = [...entities.values()].find((entity) =>
entity.entityType === "managed_resource" &&
entity.scopeKind === "company" &&
entity.scopeId === companyId &&
entity.externalId === `${manifest.id}:skill:${skillKey}`,
);
const record: PluginEntityRecord = {
id: existingEntity?.id ?? randomUUID(),
entityType: "managed_resource",
scopeKind: "company",
scopeId: companyId,
externalId: `${manifest.id}:skill:${skillKey}`,
title: declaration.displayName,
status: null,
data: { resourceKind: "skill", resourceKey: skillKey, skillId: skill.id, skill },
createdAt: existingEntity?.createdAt ?? nowIso,
updatedAt: nowIso,
};
entities.set(record.id, record);
return {
pluginKey: manifest.id,
resourceKind: "skill",
resourceKey: skillKey,
companyId,
skillId: skill.id,
skill,
status: "reset",
defaultDrift: null,
} satisfies PluginManagedSkillResolution;
},
},
},
2026-03-13 16:22:34 -05:00
companies: {
async list(input) {
requireCapability(manifest, capabilitySet, "companies.read");
let out = [...companies.values()];
if (input?.offset) out = out.slice(input.offset);
if (input?.limit) out = out.slice(0, input.limit);
return out;
},
async get(companyId) {
requireCapability(manifest, capabilitySet, "companies.read");
return companies.get(companyId) ?? null;
},
},
issues: {
async list(input) {
requireCapability(manifest, capabilitySet, "issues.read");
const companyId = requireCompanyId(input?.companyId);
let out = [...issues.values()];
out = out.filter((issue) => issue.companyId === companyId);
if (input?.projectId) out = out.filter((issue) => issue.projectId === input.projectId);
if (input?.assigneeAgentId) out = out.filter((issue) => issue.assigneeAgentId === input.assigneeAgentId);
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
if (input?.originKind) {
if (input.originKind.startsWith("plugin:")) normalizePluginOriginKind(input.originKind);
out = out.filter((issue) => issue.originKind === input.originKind);
}
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
if (input?.originKindPrefix) {
const prefix = input.originKindPrefix;
out = out.filter((issue) =>
typeof issue.originKind === "string" && issue.originKind.startsWith(prefix),
);
}
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
if (input?.originId) out = out.filter((issue) => issue.originId === input.originId);
2026-03-13 16:22:34 -05:00
if (input?.status) out = out.filter((issue) => issue.status === input.status);
if (input?.offset) out = out.slice(input.offset);
if (input?.limit) out = out.slice(0, input.limit);
return out;
},
async get(issueId, companyId) {
requireCapability(manifest, capabilitySet, "issues.read");
const issue = issues.get(issueId);
return isInCompany(issue, companyId) ? issue : null;
},
async create(input) {
requireCapability(manifest, capabilitySet, "issues.create");
const now = new Date();
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
const originKind = normalizePluginOriginKind(
input.surfaceVisibility === "plugin_operation" && !input.originKind
? pluginOperationIssueOriginKind(manifest.id)
: input.originKind,
);
2026-03-13 16:22:34 -05:00
const record: Issue = {
id: randomUUID(),
companyId: input.companyId,
projectId: input.projectId ?? null,
projectWorkspaceId: null,
2026-03-13 16:22:34 -05:00
goalId: input.goalId ?? null,
parentId: input.parentId ?? null,
title: input.title,
description: input.description ?? null,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
status: input.status ?? "todo",
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
workMode: "standard",
2026-03-13 16:22:34 -05:00
priority: input.priority ?? "medium",
assigneeAgentId: input.assigneeAgentId ?? null,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
assigneeUserId: input.assigneeUserId ?? null,
2026-03-13 16:22:34 -05:00
checkoutRunId: null,
executionRunId: null,
executionAgentNameKey: null,
executionLockedAt: null,
createdByAgentId: null,
createdByUserId: null,
issueNumber: null,
identifier: null,
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
originKind,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
originId: input.originId ?? null,
originRunId: input.originRunId ?? null,
requestDepth: input.requestDepth ?? 0,
billingCode: input.billingCode ?? null,
[codex] Add LLM Wiki plugin host support (#5597) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system needs host contracts and runtime support before large plugins can integrate cleanly. > - The source branch mixed the LLM Wiki package with supporting host/runtime work, managed plugin skills, root-level storage spaces, and a bookmarks reference plugin. > - [PAP-9173](/PAP/issues/PAP-9173) asked for the current branch to be split by file boundary: plugin package separately from everything else. > - [PAP-9188](/PAP/issues/PAP-9188) clarified that LLM Wiki may have plugin-local spaces, but Paperclip core should not reorganize top-level local storage into spaces. > - Follow-up review clarified that the bookmarks example should not ship in this PR either. > - This pull request contains the non-`packages/plugins/plugin-llm-wiki/` host/runtime work, keeps runtime state under the selected Paperclip instance root, and no longer includes the bookmarks example. ## What Changed - Added/updated plugin host contracts, SDK types, worker RPC plumbing, managed plugin skill support, and related server tests. - Removed the bookmarks example plugin package and its bundled-example/workspace references. - Removed the root-level local spaces CLI/migration surface and restored instance-root runtime defaults for config, db, logs, storage, secrets, workspaces, projects, and adapter homes. - Replaced shared root `space-paths` helpers with `home-paths` helpers for core runtime storage. - Tightened stranded recovery unique-conflict detection so concurrent recovery scans reuse the raced recovery issue when Postgres errors are wrapped. - Kept `packages/plugins/plugin-llm-wiki/` out of this PR diff; plugin-local spaces remain in the stacked plugin-only PR. ## Verification - `pnpm exec vitest run cli/src/__tests__/data-dir.test.ts cli/src/__tests__/home-paths.test.ts cli/src/__tests__/onboard.test.ts packages/shared/src/home-paths.test.ts packages/db/src/runtime-config.test.ts server/src/__tests__/agent-instructions-service.test.ts server/src/__tests__/claude-local-execute.test.ts server/src/__tests__/codex-local-execute.test.ts` - `pnpm exec vitest run packages/db/src/runtime-config.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-routes-authz.test.ts` - `pnpm --filter @paperclipai/server typecheck` - `pnpm exec vitest run server/src/__tests__/heartbeat-process-recovery.test.ts -t "reuses the raced stranded recovery issue"` skipped locally because embedded Postgres did not initialize on this macOS temp host; the code path was typechecked and is covered by Linux CI. - Boundary check: no core references remain for `PAPERCLIP_SPACE_ID`, `spaces migrate-default`, `@paperclipai/shared/space-paths`, `registerSpacesCommands`, or the removed bookmarks example. - Previous PR head `4f23e034` had green GitHub checks: `verify`, all four serialized server shards, `e2e`, `Canary Dry Run`, `policy`, Snyk, and `Greptile Review`. Current head `582f466d` is re-running checks after the bookmarks deletion. ## Risks - Plugin host changes touch shared runtime paths, so regressions would most likely appear in adapter startup, plugin loading, or local dev path defaults. - Removing the bookmarks example also removes one demonstration of plugin database namespaces plus local-folder persistence; remaining plugin examples still cover bundled example discovery and plugin host flows. - The plugin package itself is intentionally deferred to the stacked plugin-only PR, where LLM Wiki plugin-local spaces live. - Existing installs that tested the transient root-level spaces CLI should stop using it; this PR intentionally removes that unsupported migration surface before merge. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI GPT-5 Codex via Codex CLI, tool use and local code execution enabled; context window not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass, except where noted above for host-specific embedded Postgres initialization - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge Stacked follow-up: PR #5592 contains only `packages/plugins/plugin-llm-wiki/` and targets this branch. --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-10 07:34:12 -05:00
assigneeAdapterOverrides: input.assigneeAdapterOverrides ?? null,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
executionWorkspaceId: input.executionWorkspaceId ?? null,
executionWorkspacePreference: input.executionWorkspacePreference ?? null,
executionWorkspaceSettings: input.executionWorkspaceSettings ?? null,
2026-03-13 16:22:34 -05:00
startedAt: null,
completedAt: null,
cancelledAt: null,
hiddenAt: null,
createdAt: now,
updatedAt: now,
};
issues.set(record.id, record);
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
if (input.blockedByIssueIds) blockedByIssueIds.set(record.id, [...new Set(input.blockedByIssueIds)]);
2026-03-13 16:22:34 -05:00
return record;
},
async update(issueId, patch, companyId) {
requireCapability(manifest, capabilitySet, "issues.update");
const record = issues.get(issueId);
if (!isInCompany(record, companyId)) throw new Error(`Issue not found: ${issueId}`);
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
const { blockedByIssueIds: nextBlockedByIssueIds, ...issuePatch } = patch;
if (issuePatch.originKind !== undefined) {
issuePatch.originKind = normalizePluginOriginKind(issuePatch.originKind);
}
2026-03-13 16:22:34 -05:00
const updated: Issue = {
...record,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
...issuePatch,
2026-03-13 16:22:34 -05:00
updatedAt: new Date(),
};
issues.set(issueId, updated);
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
if (nextBlockedByIssueIds !== undefined) {
blockedByIssueIds.set(issueId, [...new Set(nextBlockedByIssueIds)]);
}
2026-03-13 16:22:34 -05:00
return updated;
},
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
async assertCheckoutOwner(input) {
requireCapability(manifest, capabilitySet, "issues.checkout");
const record = issues.get(input.issueId);
if (!isInCompany(record, input.companyId)) throw new Error(`Issue not found: ${input.issueId}`);
if (
record.status !== "in_progress" ||
record.assigneeAgentId !== input.actorAgentId ||
(record.checkoutRunId !== null && record.checkoutRunId !== input.actorRunId)
) {
throw new Error("Issue run ownership conflict");
}
return {
issueId: record.id,
status: record.status,
assigneeAgentId: record.assigneeAgentId,
checkoutRunId: record.checkoutRunId,
adoptedFromRunId: null,
};
},
async requestWakeup(issueId, companyId) {
requireCapability(manifest, capabilitySet, "issues.wakeup");
const record = issues.get(issueId);
if (!isInCompany(record, companyId)) throw new Error(`Issue not found: ${issueId}`);
if (!record.assigneeAgentId) throw new Error("Issue has no assigned agent to wake");
if (["backlog", "done", "cancelled"].includes(record.status)) {
throw new Error(`Issue is not wakeable in status: ${record.status}`);
}
const unresolved = issueRelationSummary(issueId).blockedBy.filter((blocker) => blocker.status !== "done");
if (unresolved.length > 0) throw new Error("Issue is blocked by unresolved blockers");
return { queued: true, runId: randomUUID() };
},
async requestWakeups(issueIds, companyId) {
requireCapability(manifest, capabilitySet, "issues.wakeup");
const results = [];
for (const issueId of issueIds) {
const record = issues.get(issueId);
if (!isInCompany(record, companyId)) throw new Error(`Issue not found: ${issueId}`);
if (!record.assigneeAgentId) throw new Error("Issue has no assigned agent to wake");
if (["backlog", "done", "cancelled"].includes(record.status)) {
throw new Error(`Issue is not wakeable in status: ${record.status}`);
}
const unresolved = issueRelationSummary(issueId).blockedBy.filter((blocker) => blocker.status !== "done");
if (unresolved.length > 0) throw new Error("Issue is blocked by unresolved blockers");
results.push({ issueId, queued: true, runId: randomUUID() });
}
return results;
},
2026-03-13 16:22:34 -05:00
async listComments(issueId, companyId) {
requireCapability(manifest, capabilitySet, "issue.comments.read");
if (!isInCompany(issues.get(issueId), companyId)) return [];
return issueComments.get(issueId) ?? [];
},
async createComment(issueId, body, companyId, options) {
2026-03-13 16:22:34 -05:00
requireCapability(manifest, capabilitySet, "issue.comments.create");
const parentIssue = issues.get(issueId);
if (!isInCompany(parentIssue, companyId)) {
throw new Error(`Issue not found: ${issueId}`);
}
const now = new Date();
const comment: IssueComment = {
id: randomUUID(),
companyId: parentIssue.companyId,
issueId,
Add recovery handoff system notices (#5289) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - Agent runs can end productively while the source issue still lacks a durable final disposition. > - That leaves the control plane unsure whether to resume, escalate, or close the work. > - Issue comments also need a presentation contract so system-authored recovery notices can render as first-class thread messages without overloading normal comments. > - This pull request adds successful-run handoff recovery, comment presentation metadata, and system notice rendering. > - The benefit is stricter task liveness with clearer operator-facing recovery state. ## What Changed - Added successful-run handoff decisions, wake payloads, escalation behavior, and recovery tests. - Added issue comment presentation metadata with migration `0078_white_darwin.sql` and shared/server/company portability support. - Rendered recovery/system notices in issue chat with dedicated UI components, fixtures, tests, and storybook/lab coverage. - Included the current recovery model-profile hint patch so automatic recovery follow-ups use the cheap profile. ## Verification - `pnpm install --frozen-lockfile` - `pnpm exec vitest run server/src/services/recovery/successful-run-handoff.test.ts ui/src/components/SystemNotice.test.tsx ui/src/lib/system-notice-comment.test.ts ui/src/components/IssueChatThreadSystemNotice.test.tsx` ## Risks - Migration-bearing PR: merge this before any other branch that might later add a migration. - The branch touches both recovery services and issue-thread rendering, so review should pay attention to recovery wake idempotency and comment metadata compatibility. ## Model Used - OpenAI GPT-5 Codex via Paperclip `codex_local` adapter, with shell/git/GitHub CLI tool use. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-06 06:05:58 -05:00
authorType: options?.authorAgentId ? "agent" : "system",
authorAgentId: options?.authorAgentId ?? null,
2026-03-13 16:22:34 -05:00
authorUserId: null,
body,
Add recovery handoff system notices (#5289) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - Agent runs can end productively while the source issue still lacks a durable final disposition. > - That leaves the control plane unsure whether to resume, escalate, or close the work. > - Issue comments also need a presentation contract so system-authored recovery notices can render as first-class thread messages without overloading normal comments. > - This pull request adds successful-run handoff recovery, comment presentation metadata, and system notice rendering. > - The benefit is stricter task liveness with clearer operator-facing recovery state. ## What Changed - Added successful-run handoff decisions, wake payloads, escalation behavior, and recovery tests. - Added issue comment presentation metadata with migration `0078_white_darwin.sql` and shared/server/company portability support. - Rendered recovery/system notices in issue chat with dedicated UI components, fixtures, tests, and storybook/lab coverage. - Included the current recovery model-profile hint patch so automatic recovery follow-ups use the cheap profile. ## Verification - `pnpm install --frozen-lockfile` - `pnpm exec vitest run server/src/services/recovery/successful-run-handoff.test.ts ui/src/components/SystemNotice.test.tsx ui/src/lib/system-notice-comment.test.ts ui/src/components/IssueChatThreadSystemNotice.test.tsx` ## Risks - Migration-bearing PR: merge this before any other branch that might later add a migration. - The branch touches both recovery services and issue-thread rendering, so review should pay attention to recovery wake idempotency and comment metadata compatibility. ## Model Used - OpenAI GPT-5 Codex via Paperclip `codex_local` adapter, with shell/git/GitHub CLI tool use. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-06 06:05:58 -05:00
presentation: null,
metadata: null,
2026-03-13 16:22:34 -05:00
createdAt: now,
updatedAt: now,
};
const current = issueComments.get(issueId) ?? [];
current.push(comment);
issueComments.set(issueId, current);
return comment;
},
[codex] Add structured issue-thread interactions (#4244) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - Operators supervise that work through issues, comments, approvals, and the board UI. > - Some agent proposals need structured board/user decisions, not hidden markdown conventions or heavyweight governed approvals. > - Issue-thread interactions already provide a natural thread-native surface for proposed tasks and questions. > - This pull request extends that surface with request confirmations, richer interaction cards, and agent/plugin/MCP helpers. > - The benefit is that plan approvals and yes/no decisions become explicit, auditable, and resumable without losing the single-issue workflow. ## What Changed - Added persisted issue-thread interactions for suggested tasks, structured questions, and request confirmations. - Added board UI cards for interaction review, selection, question answers, and accept/reject confirmation flows. - Added MCP and plugin SDK helpers for creating interaction cards from agents/plugins. - Updated agent wake instructions, onboarding assets, Paperclip skill docs, and public docs to prefer structured confirmations for issue-scoped decisions. - Rebased the branch onto `public-gh/master` and renumbered branch migrations to `0063` and `0064`; the idempotency migration uses `ADD COLUMN IF NOT EXISTS` for old branch users. ## Verification - `git diff --check public-gh/master..HEAD` - `pnpm exec vitest run packages/adapter-utils/src/server-utils.test.ts packages/mcp-server/src/tools.test.ts packages/shared/src/issue-thread-interactions.test.ts ui/src/lib/issue-thread-interactions.test.ts ui/src/lib/issue-chat-messages.test.ts ui/src/components/IssueThreadInteractionCard.test.tsx ui/src/components/IssueChatThread.test.tsx server/src/__tests__/issue-thread-interaction-routes.test.ts server/src/__tests__/issue-thread-interactions-service.test.ts server/src/services/issue-thread-interactions.test.ts` -> 9 files / 79 tests passed - `pnpm -r typecheck` -> passed, including `packages/db` migration numbering check ## Risks - Medium: this adds a new issue-thread interaction model across db/shared/server/ui/plugin surfaces. - Migration risk is reduced by placing this branch after current master migrations (`0063`, `0064`) and making the idempotency column add idempotent for users who applied the old branch numbering. - UI interaction behavior is covered by component tests, but this PR does not include browser screenshots. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5-class coding agent runtime. Exact model ID and context window are not exposed in this Paperclip run; tool use and local shell/code execution were enabled. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-21 20:15:11 -05:00
async createInteraction(issueId, interaction, companyId, options) {
requireCapability(manifest, capabilitySet, "issue.interactions.create");
const parentIssue = issues.get(issueId);
if (!isInCompany(parentIssue, companyId)) {
throw new Error(`Issue not found: ${issueId}`);
}
const now = new Date();
const current = issueInteractions.get(issueId) ?? [];
if (interaction.idempotencyKey) {
const existing = current.find((entry) => entry.idempotencyKey === interaction.idempotencyKey);
if (existing) return existing;
}
const created: IssueThreadInteraction = {
id: randomUUID(),
companyId: parentIssue.companyId,
issueId,
kind: interaction.kind,
status: "pending",
continuationPolicy: interaction.continuationPolicy ?? "wake_assignee",
idempotencyKey: interaction.idempotencyKey ?? null,
sourceCommentId: interaction.sourceCommentId ?? null,
sourceRunId: interaction.sourceRunId ?? null,
title: interaction.title ?? null,
summary: interaction.summary ?? null,
createdByAgentId: options?.authorAgentId ?? null,
createdByUserId: null,
payload: interaction.payload,
result: null,
createdAt: now,
updatedAt: now,
} as IssueThreadInteraction;
current.push(created);
issueInteractions.set(issueId, current);
return created;
},
async suggestTasks(issueId, interaction, companyId, options) {
return this.createInteraction(issueId, { ...interaction, kind: "suggest_tasks" }, companyId, options) as Promise<any>;
},
async askUserQuestions(issueId, interaction, companyId, options) {
return this.createInteraction(issueId, { ...interaction, kind: "ask_user_questions" }, companyId, options) as Promise<any>;
},
async requestConfirmation(issueId, interaction, companyId, options) {
return this.createInteraction(issueId, { ...interaction, kind: "request_confirmation" }, companyId, options) as Promise<any>;
},
documents: {
async list(issueId, companyId) {
requireCapability(manifest, capabilitySet, "issue.documents.read");
if (!isInCompany(issues.get(issueId), companyId)) return [];
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
return [...issueDocuments.values()]
.filter((document) => document.issueId === issueId && document.companyId === companyId)
.map(({ body: _body, ...summary }) => summary);
},
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
async get(issueId, key, companyId) {
requireCapability(manifest, capabilitySet, "issue.documents.read");
if (!isInCompany(issues.get(issueId), companyId)) return null;
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
return issueDocuments.get(`${issueId}|${key}`) ?? null;
},
async upsert(input) {
requireCapability(manifest, capabilitySet, "issue.documents.write");
const parentIssue = issues.get(input.issueId);
if (!isInCompany(parentIssue, input.companyId)) {
throw new Error(`Issue not found: ${input.issueId}`);
}
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
const now = new Date();
const existing = issueDocuments.get(`${input.issueId}|${input.key}`);
const document: IssueDocument = {
id: existing?.id ?? randomUUID(),
companyId: input.companyId,
issueId: input.issueId,
key: input.key,
title: input.title ?? existing?.title ?? null,
format: "markdown",
latestRevisionId: randomUUID(),
latestRevisionNumber: (existing?.latestRevisionNumber ?? 0) + 1,
createdByAgentId: existing?.createdByAgentId ?? null,
createdByUserId: existing?.createdByUserId ?? null,
updatedByAgentId: null,
updatedByUserId: null,
[codex] Add issue document locking (#6009) ## Thinking Path > - Paperclip orchestrates AI-agent companies through company-scoped issues, comments, and issue documents. > - Issue documents are the durable place where plans, handoffs, and other work artifacts are revised over time. > - Some documents need to be preserved as operator-approved snapshots while agents continue working on the same issue. > - Without document locking, a later board or agent write can overwrite the document key that reviewers expected to remain stable. > - This pull request adds board-managed issue document locks and makes agent writes to locked keys create a derived document instead of mutating the locked document. > - The benefit is safer document handoffs: approved or frozen issue documents stay immutable until the board explicitly unlocks them. ## What Changed - Added `locked_at`, `locked_by_agent_id`, and `locked_by_user_id` document fields plus migration `0085_tranquil_the_executioner.sql`. - Added document lock/unlock service behavior, route endpoints, activity events, and locked-document write protections. - Made agent document writes to locked keys create a new derived key such as `plan-2` rather than overwriting the locked document. - Surfaced lock state through shared issue document types, UI API methods, document header lock controls, and activity formatting. - Added server and UI tests for lock/unlock behavior, locked document immutability, and UI action visibility. - Updated `doc/SPEC-implementation.md` with the V1 document lock contract and endpoints. ## Verification - `git rebase public-gh/master` completed cleanly after committing the branch changes. - `git diff --check` passed before commit. - `pnpm run preflight:workspace-links && pnpm exec vitest run server/src/__tests__/documents-service.test.ts server/src/__tests__/issue-agent-mutation-ownership-routes.test.ts ui/src/components/IssueDocumentsSection.test.tsx ui/src/components/IssueContinuationHandoff.test.tsx ui/src/lib/document-revisions.test.ts` passed: 5 files, 32 tests. ## Risks - Medium risk because this changes the document persistence contract and adds a migration. - The migration uses `ADD COLUMN IF NOT EXISTS` and guarded foreign-key creation so it remains safe for users who may have already applied an earlier copy of the migration. - Locked documents intentionally reject board edits/deletes/restores until unlocked; any existing workflows that expected direct overwrite need to unlock first. - Agent writes to locked keys now create derived documents, which may create extra issue documents when agents retry locked writes. ## Model Used - OpenAI Codex coding agent based on GPT-5, with tool use and local code execution in the Paperclip worktree. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-15 08:54:55 -05:00
lockedAt: existing?.lockedAt ?? null,
lockedByAgentId: existing?.lockedByAgentId ?? null,
lockedByUserId: existing?.lockedByUserId ?? null,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
createdAt: existing?.createdAt ?? now,
updatedAt: now,
body: input.body,
};
issueDocuments.set(`${input.issueId}|${input.key}`, document);
return document;
},
async delete(issueId, _key, companyId) {
requireCapability(manifest, capabilitySet, "issue.documents.write");
const parentIssue = issues.get(issueId);
if (!isInCompany(parentIssue, companyId)) {
throw new Error(`Issue not found: ${issueId}`);
}
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
issueDocuments.delete(`${issueId}|${_key}`);
},
},
relations: {
async get(issueId, companyId) {
requireCapability(manifest, capabilitySet, "issue.relations.read");
if (!isInCompany(issues.get(issueId), companyId)) throw new Error(`Issue not found: ${issueId}`);
return issueRelationSummary(issueId);
},
async setBlockedBy(issueId, nextBlockedByIssueIds, companyId) {
requireCapability(manifest, capabilitySet, "issue.relations.write");
if (!isInCompany(issues.get(issueId), companyId)) throw new Error(`Issue not found: ${issueId}`);
blockedByIssueIds.set(issueId, [...new Set(nextBlockedByIssueIds)]);
return issueRelationSummary(issueId);
},
async addBlockers(issueId, blockerIssueIds, companyId) {
requireCapability(manifest, capabilitySet, "issue.relations.write");
if (!isInCompany(issues.get(issueId), companyId)) throw new Error(`Issue not found: ${issueId}`);
const next = new Set(blockedByIssueIds.get(issueId) ?? []);
for (const blockerIssueId of blockerIssueIds) next.add(blockerIssueId);
blockedByIssueIds.set(issueId, [...next]);
return issueRelationSummary(issueId);
},
async removeBlockers(issueId, blockerIssueIds, companyId) {
requireCapability(manifest, capabilitySet, "issue.relations.write");
if (!isInCompany(issues.get(issueId), companyId)) throw new Error(`Issue not found: ${issueId}`);
const removals = new Set(blockerIssueIds);
blockedByIssueIds.set(
issueId,
(blockedByIssueIds.get(issueId) ?? []).filter((blockerIssueId) => !removals.has(blockerIssueId)),
);
return issueRelationSummary(issueId);
},
},
async getSubtree(issueId, companyId, options) {
requireCapability(manifest, capabilitySet, "issue.subtree.read");
const root = issues.get(issueId);
if (!isInCompany(root, companyId)) throw new Error(`Issue not found: ${issueId}`);
const includeRoot = options?.includeRoot !== false;
const allIds = [root.id];
let frontier = [root.id];
while (frontier.length > 0) {
const children = [...issues.values()]
.filter((issue) => issue.companyId === companyId && frontier.includes(issue.parentId ?? ""))
.map((issue) => issue.id)
.filter((id) => !allIds.includes(id));
allIds.push(...children);
frontier = children;
}
const issueIds = includeRoot ? allIds : allIds.filter((id) => id !== root.id);
const subtreeIssues = issueIds.map((id) => issues.get(id)).filter((candidate): candidate is Issue => Boolean(candidate));
return {
rootIssueId: root.id,
companyId,
issueIds,
issues: subtreeIssues,
...(options?.includeRelations
? { relations: Object.fromEntries(issueIds.map((id) => [id, issueRelationSummary(id)])) }
: {}),
...(options?.includeDocuments ? { documents: Object.fromEntries(issueIds.map((id) => [id, []])) } : {}),
...(options?.includeActiveRuns ? { activeRuns: Object.fromEntries(issueIds.map((id) => [id, []])) } : {}),
...(options?.includeAssignees ? { assignees: {} } : {}),
};
},
summaries: {
async getOrchestration(input) {
requireCapability(manifest, capabilitySet, "issues.orchestration.read");
const root = issues.get(input.issueId);
if (!isInCompany(root, input.companyId)) throw new Error(`Issue not found: ${input.issueId}`);
const subtreeIssueIds = [root.id];
if (input.includeSubtree) {
let frontier = [root.id];
while (frontier.length > 0) {
const children = [...issues.values()]
.filter((issue) => issue.companyId === input.companyId && frontier.includes(issue.parentId ?? ""))
.map((issue) => issue.id)
.filter((id) => !subtreeIssueIds.includes(id));
subtreeIssueIds.push(...children);
frontier = children;
}
}
return {
issueId: root.id,
companyId: input.companyId,
subtreeIssueIds,
relations: Object.fromEntries(subtreeIssueIds.map((id) => [id, issueRelationSummary(id)])),
approvals: [],
runs: [],
costs: {
costCents: 0,
inputTokens: 0,
cachedInputTokens: 0,
outputTokens: 0,
billingCode: input.billingCode ?? null,
},
openBudgetIncidents: [],
invocationBlocks: [],
};
},
},
2026-03-13 16:22:34 -05:00
},
agents: {
async list(input) {
requireCapability(manifest, capabilitySet, "agents.read");
const companyId = requireCompanyId(input?.companyId);
let out = [...agents.values()];
out = out.filter((agent) => agent.companyId === companyId);
if (input?.status) out = out.filter((agent) => agent.status === input.status);
if (input?.offset) out = out.slice(input.offset);
if (input?.limit) out = out.slice(0, input.limit);
return out;
},
async get(agentId, companyId) {
requireCapability(manifest, capabilitySet, "agents.read");
const agent = agents.get(agentId);
return isInCompany(agent, companyId) ? agent : null;
},
async pause(agentId, companyId) {
requireCapability(manifest, capabilitySet, "agents.pause");
const cid = requireCompanyId(companyId);
const agent = agents.get(agentId);
if (!isInCompany(agent, cid)) throw new Error(`Agent not found: ${agentId}`);
if (agent!.status === "terminated") throw new Error("Cannot pause terminated agent");
const updated: Agent = { ...agent!, status: "paused", updatedAt: new Date() };
agents.set(agentId, updated);
return updated;
},
async resume(agentId, companyId) {
requireCapability(manifest, capabilitySet, "agents.resume");
const cid = requireCompanyId(companyId);
const agent = agents.get(agentId);
if (!isInCompany(agent, cid)) throw new Error(`Agent not found: ${agentId}`);
if (agent!.status === "terminated") throw new Error("Cannot resume terminated agent");
if (agent!.status === "pending_approval") throw new Error("Pending approval agents cannot be resumed");
const updated: Agent = { ...agent!, status: "idle", updatedAt: new Date() };
agents.set(agentId, updated);
return updated;
},
async invoke(agentId, companyId, opts) {
requireCapability(manifest, capabilitySet, "agents.invoke");
const cid = requireCompanyId(companyId);
const agent = agents.get(agentId);
if (!isInCompany(agent, cid)) throw new Error(`Agent not found: ${agentId}`);
if (
agent!.status === "paused" ||
agent!.status === "terminated" ||
agent!.status === "pending_approval"
) {
throw new Error(`Agent is not invokable in its current state: ${agent!.status}`);
}
return { runId: randomUUID() };
},
Expand plugin host surface (#5205) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The plugin system is the extension boundary for optional product capabilities > - Rich plugins need more than a worker entrypoint: they need scoped database storage, local project folders, managed agents/routines, host navigation, and reusable UI components > - The LLM Wiki work exposed those missing host surfaces while keeping plugin code outside the core control plane > - This pull request expands the core plugin host, SDK, server APIs, and UI bridge so plugins can declare and use those surfaces > - The benefit is that future plugins can integrate with Paperclip through documented, validated contracts instead of bespoke server or UI imports ## What Changed - Added plugin-managed database namespaces and migration tracking, including Drizzle schema/migration files and SQL validation for namespace isolation. - Added server support for plugin local folders, managed agents, managed routines, scoped plugin APIs, and plugin operation visibility. - Expanded shared plugin manifest/types/validators and SDK host/testing/UI exports for richer plugin surfaces. - Added reusable UI pieces for file trees, managed routines, resizable sidebars, route sidebars, and plugin bridge initialization. - Updated plugin docs and example plugins to use the expanded host and SDK surface. ## Verification - `pnpm install --frozen-lockfile` - `pnpm run preflight:workspace-links && pnpm exec vitest run packages/shared/src/validators/plugin.test.ts server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-local-folders.test.ts server/src/__tests__/plugin-managed-agents.test.ts server/src/__tests__/plugin-managed-routines.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts ui/src/api/plugins.test.ts ui/src/components/FileTree.test.tsx ui/src/components/ResizableSidebarPane.test.tsx ui/src/pages/PluginPage.test.tsx ui/src/plugins/bridge.test.ts` passed: 11 files, 67 tests. - Confirmed this PR changes 89 files and does not include `pnpm-lock.yaml` or `.github/workflows/*`. ## Risks - Medium: this expands plugin host contracts across db/shared/server/ui and includes a new core migration (`0076_useful_elektra.sql`). - The plugin database namespace validator is intentionally restrictive; plugin authors may need follow-up affordances for SQL patterns that remain blocked. - Merge this before the LLM Wiki plugin PR so the plugin can resolve the new SDK and host APIs. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used - OpenAI Codex, GPT-5 coding agent, tool-enabled shell/git/GitHub workflow. Context window size was not exposed by the runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-05 07:42:57 -05:00
managed: {
async get(agentKey, companyId) {
requireCapability(manifest, capabilitySet, "agents.managed");
const cid = requireCompanyId(companyId);
managedAgentDeclaration(agentKey);
const agent = [...agents.values()].find((candidate) =>
candidate.companyId === cid &&
candidate.status !== "terminated" &&
isManagedAgent(candidate, agentKey),
) ?? null;
return managedResolution(agentKey, cid, agent, agent ? "resolved" : "missing");
},
async reconcile(agentKey, companyId) {
requireCapability(manifest, capabilitySet, "agents.managed");
const cid = requireCompanyId(companyId);
const declaration = managedAgentDeclaration(agentKey);
const existingAgent = [...agents.values()].find((candidate) =>
candidate.companyId === cid &&
candidate.status !== "terminated" &&
isManagedAgent(candidate, agentKey),
) ?? null;
const existing = managedResolution(agentKey, cid, existingAgent, existingAgent ? "resolved" : "missing");
if (existing.agent) return existing;
const now = new Date();
const created: Agent = {
id: randomUUID(),
companyId: cid,
name: declaration.displayName,
urlKey: declaration.displayName.toLowerCase().replace(/[^a-z0-9]+/g, "-").replace(/^-|-$/g, ""),
role: (declaration.role ?? "general") as Agent["role"],
title: declaration.title ?? null,
icon: declaration.icon ?? null,
status: declaration.status ?? "idle",
reportsTo: null,
capabilities: declaration.capabilities ?? null,
adapterType: (declaration.adapterType ?? "process") as Agent["adapterType"],
adapterConfig: declaration.adapterConfig ?? {},
runtimeConfig: declaration.runtimeConfig ?? {},
budgetMonthlyCents: declaration.budgetMonthlyCents ?? 0,
spentMonthlyCents: 0,
pauseReason: null,
pausedAt: null,
permissions: { canCreateAgents: Boolean(declaration.permissions?.canCreateAgents) },
lastHeartbeatAt: null,
metadata: managedAgentMetadata(agentKey),
createdAt: now,
updatedAt: now,
};
agents.set(created.id, created);
return managedResolution(agentKey, cid, created, "created");
},
async reset(agentKey, companyId) {
requireCapability(manifest, capabilitySet, "agents.managed");
const cid = requireCompanyId(companyId);
const declaration = managedAgentDeclaration(agentKey);
let agent = [...agents.values()].find((candidate) =>
candidate.companyId === cid &&
candidate.status !== "terminated" &&
isManagedAgent(candidate, agentKey),
) ?? null;
if (!agent) {
const now = new Date();
agent = {
id: randomUUID(),
companyId: cid,
name: declaration.displayName,
urlKey: declaration.displayName.toLowerCase().replace(/[^a-z0-9]+/g, "-").replace(/^-|-$/g, ""),
role: (declaration.role ?? "general") as Agent["role"],
title: declaration.title ?? null,
icon: declaration.icon ?? null,
status: declaration.status ?? "idle",
reportsTo: null,
capabilities: declaration.capabilities ?? null,
adapterType: (declaration.adapterType ?? "process") as Agent["adapterType"],
adapterConfig: declaration.adapterConfig ?? {},
runtimeConfig: declaration.runtimeConfig ?? {},
budgetMonthlyCents: declaration.budgetMonthlyCents ?? 0,
spentMonthlyCents: 0,
pauseReason: null,
pausedAt: null,
permissions: { canCreateAgents: Boolean(declaration.permissions?.canCreateAgents) },
lastHeartbeatAt: null,
metadata: managedAgentMetadata(agentKey),
createdAt: now,
updatedAt: now,
};
agents.set(agent.id, agent);
}
const resolved = managedResolution(agentKey, cid, agent, "resolved");
if (!resolved.agent) return resolved;
const updated: Agent = {
...resolved.agent,
name: declaration.displayName,
role: (declaration.role ?? "general") as Agent["role"],
title: declaration.title ?? null,
icon: declaration.icon ?? null,
capabilities: declaration.capabilities ?? null,
adapterType: (declaration.adapterType ?? "process") as Agent["adapterType"],
adapterConfig: declaration.adapterConfig ?? {},
runtimeConfig: declaration.runtimeConfig ?? {},
budgetMonthlyCents: declaration.budgetMonthlyCents ?? 0,
permissions: { canCreateAgents: Boolean(declaration.permissions?.canCreateAgents) },
metadata: managedAgentMetadata(agentKey, resolved.agent.metadata),
updatedAt: new Date(),
};
agents.set(updated.id, updated);
return managedResolution(agentKey, cid, updated, "reset");
},
},
2026-03-13 16:22:34 -05:00
sessions: {
async create(agentId, companyId, opts) {
requireCapability(manifest, capabilitySet, "agent.sessions.create");
const cid = requireCompanyId(companyId);
const agent = agents.get(agentId);
if (!isInCompany(agent, cid)) throw new Error(`Agent not found: ${agentId}`);
const session: AgentSession = {
sessionId: randomUUID(),
agentId,
companyId: cid,
status: "active",
createdAt: new Date().toISOString(),
};
sessions.set(session.sessionId, session);
return session;
},
async list(agentId, companyId) {
requireCapability(manifest, capabilitySet, "agent.sessions.list");
const cid = requireCompanyId(companyId);
return [...sessions.values()].filter(
(s) => s.agentId === agentId && s.companyId === cid && s.status === "active",
);
},
async sendMessage(sessionId, companyId, opts) {
requireCapability(manifest, capabilitySet, "agent.sessions.send");
const session = sessions.get(sessionId);
if (!session || session.status !== "active") throw new Error(`Session not found or closed: ${sessionId}`);
if (session.companyId !== companyId) throw new Error(`Session not found: ${sessionId}`);
if (opts.onEvent) {
sessionEventCallbacks.set(sessionId, opts.onEvent);
}
return { runId: randomUUID() };
},
async close(sessionId, companyId) {
requireCapability(manifest, capabilitySet, "agent.sessions.close");
const session = sessions.get(sessionId);
if (!session) throw new Error(`Session not found: ${sessionId}`);
if (session.companyId !== companyId) throw new Error(`Session not found: ${sessionId}`);
session.status = "closed";
sessionEventCallbacks.delete(sessionId);
},
},
},
goals: {
async list(input) {
requireCapability(manifest, capabilitySet, "goals.read");
const companyId = requireCompanyId(input?.companyId);
let out = [...goals.values()];
out = out.filter((goal) => goal.companyId === companyId);
if (input?.level) out = out.filter((goal) => goal.level === input.level);
if (input?.status) out = out.filter((goal) => goal.status === input.status);
if (input?.offset) out = out.slice(input.offset);
if (input?.limit) out = out.slice(0, input.limit);
return out;
},
async get(goalId, companyId) {
requireCapability(manifest, capabilitySet, "goals.read");
const goal = goals.get(goalId);
return isInCompany(goal, companyId) ? goal : null;
},
async create(input) {
requireCapability(manifest, capabilitySet, "goals.create");
const now = new Date();
const record: Goal = {
id: randomUUID(),
companyId: input.companyId,
title: input.title,
description: input.description ?? null,
level: input.level ?? "task",
status: input.status ?? "planned",
parentId: input.parentId ?? null,
ownerAgentId: input.ownerAgentId ?? null,
createdAt: now,
updatedAt: now,
};
goals.set(record.id, record);
return record;
},
async update(goalId, patch, companyId) {
requireCapability(manifest, capabilitySet, "goals.update");
const record = goals.get(goalId);
if (!isInCompany(record, companyId)) throw new Error(`Goal not found: ${goalId}`);
const updated: Goal = {
...record,
...patch,
updatedAt: new Date(),
};
goals.set(goalId, updated);
return updated;
},
},
data: {
register(key, handler) {
dataHandlers.set(key, handler);
},
},
actions: {
register(key, handler) {
actionHandlers.set(key, handler);
},
},
streams: (() => {
const channelCompanyMap = new Map<string, string>();
return {
open(channel: string, companyId: string) {
channelCompanyMap.set(channel, companyId);
},
emit(_channel: string, _event: unknown) {
// No-op in test harness — events are not forwarded
},
close(channel: string) {
channelCompanyMap.delete(channel);
},
};
})(),
tools: {
register(name, _decl, fn) {
requireCapability(manifest, capabilitySet, "agent.tools.register");
toolHandlers.set(name, fn);
},
},
metrics: {
async write(name, value, tags) {
requireCapability(manifest, capabilitySet, "metrics.write");
metrics.push({ name, value, tags });
},
},
telemetry: {
async track(eventName, dimensions) {
requireCapability(manifest, capabilitySet, "telemetry.track");
telemetry.push({ eventName, dimensions });
},
},
2026-03-13 16:22:34 -05:00
logger: {
info(message, meta) {
logs.push({ level: "info", message, meta });
},
warn(message, meta) {
logs.push({ level: "warn", message, meta });
},
error(message, meta) {
logs.push({ level: "error", message, meta });
},
debug(message, meta) {
logs.push({ level: "debug", message, meta });
},
},
};
const harness: TestHarness = {
ctx,
seed(input) {
for (const row of input.companies ?? []) companies.set(row.id, row);
for (const row of input.projects ?? []) projects.set(row.id, row);
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
for (const row of input.issues ?? []) {
issues.set(row.id, row);
if (row.blockedBy) {
blockedByIssueIds.set(row.id, row.blockedBy.map((blocker) => blocker.id));
}
}
2026-03-13 16:22:34 -05:00
for (const row of input.issueComments ?? []) {
const list = issueComments.get(row.issueId) ?? [];
list.push(row);
issueComments.set(row.issueId, list);
}
for (const row of input.agents ?? []) agents.set(row.id, row);
for (const row of input.goals ?? []) goals.set(row.id, row);
},
setConfig(config) {
currentConfig = { ...config };
},
async emit(eventType, payload, base) {
const event: PluginEvent = {
eventId: base?.eventId ?? randomUUID(),
eventType,
companyId: base?.companyId ?? "test-company",
occurredAt: base?.occurredAt ?? new Date().toISOString(),
actorId: base?.actorId,
actorType: base?.actorType,
entityId: base?.entityId,
entityType: base?.entityType,
payload,
};
for (const handler of events) {
const exactMatch = handler.name === event.eventType;
const wildcardPluginAll = handler.name === "plugin.*" && String(event.eventType).startsWith("plugin.");
const wildcardPluginOne = String(handler.name).endsWith(".*")
&& String(event.eventType).startsWith(String(handler.name).slice(0, -1));
if (!exactMatch && !wildcardPluginAll && !wildcardPluginOne) continue;
if (!allowsEvent(handler.filter, event)) continue;
await handler.fn(event);
}
},
async runJob(jobKey, partial = {}) {
const handler = jobs.get(jobKey);
if (!handler) throw new Error(`No job handler registered for '${jobKey}'`);
await handler({
jobKey,
runId: partial.runId ?? randomUUID(),
trigger: partial.trigger ?? "manual",
scheduledAt: partial.scheduledAt ?? new Date().toISOString(),
});
},
async getData<T = unknown>(key: string, params: Record<string, unknown> = {}) {
const handler = dataHandlers.get(key);
if (!handler) throw new Error(`No data handler registered for '${key}'`);
return await handler(params) as T;
},
async performAction<T = unknown>(key: string, params: Record<string, unknown> = {}) {
const handler = actionHandlers.get(key);
if (!handler) throw new Error(`No action handler registered for '${key}'`);
return await handler(params) as T;
},
async executeTool<T = ToolResult>(name: string, params: unknown, runCtx: Partial<ToolRunContext> = {}) {
const handler = toolHandlers.get(name);
if (!handler) throw new Error(`No tool handler registered for '${name}'`);
const ctxToPass: ToolRunContext = {
agentId: runCtx.agentId ?? "agent-test",
runId: runCtx.runId ?? randomUUID(),
companyId: runCtx.companyId ?? "company-test",
projectId: runCtx.projectId ?? "project-test",
};
return await handler(params, ctxToPass) as T;
},
getState(input) {
return state.get(stateMapKey(input));
},
simulateSessionEvent(sessionId, event) {
const cb = sessionEventCallbacks.get(sessionId);
if (!cb) throw new Error(`No active session event callback for session: ${sessionId}`);
cb({ ...event, sessionId });
},
logs,
activity,
metrics,
telemetry,
[codex] Add plugin orchestration host APIs (#4114) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The plugin system is the extension path for optional capabilities that should not require core product changes for every integration. > - Plugins need scoped host APIs for issue orchestration, documents, wakeups, summaries, activity attribution, and isolated database state. > - Without those host APIs, richer plugins either cannot coordinate Paperclip work safely or need privileged core-side special cases. > - This pull request adds the plugin orchestration host surface, scoped route dispatch, a database namespace layer, and a smoke plugin that exercises the contract. > - The benefit is a broader plugin API that remains company-scoped, auditable, and covered by tests. ## What Changed - Added plugin orchestration host APIs for issue creation, document access, wakeups, summaries, plugin-origin activity, and scoped API route dispatch. - Added plugin database namespace tables, schema exports, migration checks, and idempotent replay coverage under migration `0059_plugin_database_namespaces`. - Added shared plugin route/API types and validators used by server and SDK boundaries. - Expanded plugin SDK types, protocol helpers, worker RPC host behavior, and testing utilities for orchestration flows. - Added the `plugin-orchestration-smoke-example` package to exercise scoped routes, restricted database namespaces, issue orchestration, documents, wakeups, summaries, and UI status surfaces. - Kept the new orchestration smoke fixture out of the root pnpm workspace importer so this PR preserves the repository policy of not committing `pnpm-lock.yaml`. - Updated plugin docs and database docs for the new orchestration and database namespace surfaces. - Rebased the branch onto `public-gh/master`, resolved conflicts, and removed `pnpm-lock.yaml` from the final PR diff. ## Verification - `pnpm install --frozen-lockfile` - `pnpm --filter @paperclipai/db typecheck` - `pnpm exec vitest run packages/db/src/client.test.ts` - `pnpm exec vitest run server/src/__tests__/plugin-database.test.ts server/src/__tests__/plugin-orchestration-apis.test.ts server/src/__tests__/plugin-routes-authz.test.ts server/src/__tests__/plugin-scoped-api-routes.test.ts server/src/__tests__/plugin-sdk-orchestration-contract.test.ts` - From `packages/plugins/examples/plugin-orchestration-smoke-example`: `pnpm exec vitest run --config ./vitest.config.ts` - `pnpm --dir packages/plugins/examples/plugin-orchestration-smoke-example run typecheck` - `pnpm --filter @paperclipai/server typecheck` - PR CI on latest head `293fc67c`: `policy`, `verify`, `e2e`, and `security/snyk` all passed. ## Risks - Medium risk: this expands plugin host authority, so route auth, company scoping, and plugin-origin activity attribution need careful review. - Medium risk: database namespace migration behavior must remain idempotent for environments that may have seen earlier branch versions. - Medium risk: the orchestration smoke fixture is intentionally excluded from the root workspace importer to avoid a `pnpm-lock.yaml` PR diff; direct fixture verification remains listed above. - Low operational risk from the PR setup itself: the branch is rebased onto current `master`, the migration is ordered after upstream `0057`/`0058`, and `pnpm-lock.yaml` is not in the final diff. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. Roadmap checked: this work aligns with the completed Plugin system milestone and extends the plugin surface rather than duplicating an unrelated planned core feature. ## Model Used - OpenAI Codex, GPT-5-based coding agent in a tool-enabled CLI environment. Exact hosted model build and context-window size are not exposed by the runtime; reasoning/tool use were enabled for repository inspection, editing, testing, git operations, and PR creation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] If this change affects the UI, I have included before/after screenshots (N/A: no core UI screen change; example plugin UI contract is covered by tests) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-04-20 08:52:51 -05:00
dbQueries,
dbExecutes,
2026-03-13 16:22:34 -05:00
};
return harness;
}
Add sandbox environment support (#4415) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies. > - The environment/runtime layer decides where agent work executes and how the control plane reaches those runtimes. > - Today Paperclip can run locally and over SSH, but sandboxed execution needs a first-class environment model instead of one-off adapter behavior. > - We also want sandbox providers to be pluggable so the core does not hardcode every provider implementation. > - This branch adds the Sandbox environment path, the provider contract, and a deterministic fake provider plugin. > - That required synchronized changes across shared contracts, plugin SDK surfaces, server runtime orchestration, and the UI environment/workspace flows. > - The result is that sandbox execution becomes a core control-plane capability while keeping provider implementations extensible and testable. ## What Changed - Added sandbox runtime support to the environment execution path, including runtime URL discovery, sandbox execution targeting, orchestration, and heartbeat integration. - Added plugin-provider support for sandbox environments so providers can be supplied via plugins instead of hardcoded server logic. - Added the fake sandbox provider plugin with deterministic behavior suitable for local and automated testing. - Updated shared types, validators, plugin protocol definitions, and SDK helpers to carry sandbox provider and workspace-runtime contracts across package boundaries. - Updated server routes and services so companies can create sandbox environments, select them for work, and execute work through the sandbox runtime path. - Updated the UI environment and workspace surfaces to expose sandbox environment configuration and selection. - Added test coverage for sandbox runtime behavior, provider seams, environment route guards, orchestration, and the fake provider plugin. ## Verification - Ran locally before the final fixture-only scrub: - `pnpm -r typecheck` - `pnpm test:run` - `pnpm build` - Ran locally after the final scrub amend: - `pnpm vitest run server/src/__tests__/runtime-api.test.ts` - Reviewer spot checks: - create a sandbox environment backed by the fake provider plugin - run work through that environment - confirm sandbox provider execution does not inherit host secrets implicitly ## Risks - This touches shared contracts, plugin SDK plumbing, server runtime orchestration, and UI environment/workspace flows, so regressions would likely show up as cross-layer mismatches rather than isolated type errors. - Runtime URL discovery and sandbox callback selection are sensitive to host/bind configuration; if that logic is wrong, sandbox-backed callbacks may fail even when execution succeeds. - The fake provider plugin is intentionally deterministic and test-oriented; future providers may expose capability gaps that this branch does not yet cover. ## Model Used - OpenAI Codex coding agent on a GPT-5-class backend in the Paperclip/Codex harness. Exact backend model ID is not exposed in-session. Tool-assisted workflow with shell execution, file editing, git history inspection, and local test execution. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] If this change affects the UI, I have included before/after screenshots - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] I will address all Greptile and reviewer comments before requesting merge
2026-04-24 12:15:53 -07:00
/**
* Create an environment-aware test harness that wraps the base harness with
* environment driver simulation and lifecycle event recording.
*
* Use this to test environment plugins through the full host contract:
* validateConfig probe acquireLease realizeWorkspace execute releaseLease.
*/
export function createEnvironmentTestHarness(options: EnvironmentTestHarnessOptions): EnvironmentTestHarness {
const base = createTestHarness(options);
const environmentEvents: EnvironmentEventRecord[] = [];
const driver = options.environmentDriver;
function record(
type: EnvironmentEventRecord["type"],
params: Record<string, unknown>,
result?: unknown,
error?: string,
): EnvironmentEventRecord {
const event: EnvironmentEventRecord = {
type,
driverKey: (params as { driverKey?: string }).driverKey ?? driver.driverKey,
environmentId: (params as { environmentId?: string }).environmentId ?? "unknown",
timestamp: new Date().toISOString(),
params,
result,
error,
};
environmentEvents.push(event);
return event;
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
async function callHook<R>(
type: EnvironmentEventRecord["type"],
hook: ((...args: any[]) => Promise<R>) | undefined,
params: unknown,
hookName: string,
): Promise<R> {
if (!hook) {
const err = `Environment driver '${driver.driverKey}' does not implement ${hookName}`;
record(type, params as Record<string, unknown>, undefined, err);
throw new Error(err);
}
try {
const result = await hook(params);
record(type, params as Record<string, unknown>, result);
return result;
} catch (e) {
const msg = e instanceof Error ? e.message : String(e);
record(type, params as Record<string, unknown>, undefined, msg);
throw e;
}
}
const envHarness: EnvironmentTestHarness = {
...base,
environmentEvents,
async validateConfig(params) {
return callHook("validateConfig", driver.onValidateConfig, params, "onValidateConfig");
},
async probe(params) {
return callHook("probe", driver.onProbe, params, "onProbe");
},
async acquireLease(params) {
return callHook("acquireLease", driver.onAcquireLease, params, "onAcquireLease");
},
async resumeLease(params) {
return callHook("resumeLease", driver.onResumeLease, params, "onResumeLease");
},
async releaseLease(params) {
return callHook("releaseLease", driver.onReleaseLease, params, "onReleaseLease");
},
async destroyLease(params) {
return callHook("destroyLease", driver.onDestroyLease, params, "onDestroyLease");
},
async realizeWorkspace(params) {
return callHook("realizeWorkspace", driver.onRealizeWorkspace, params, "onRealizeWorkspace");
},
async execute(params) {
return callHook("execute", driver.onExecute, params, "onExecute");
},
};
return envHarness;
}