Harden control-plane safety and issue identifiers (#5292)

## Thinking Path

> - Paperclip relies on issue identifiers, execution policies, and agent
heartbeat rules to keep autonomous work auditable.
> - Safety checks need to reject ambiguous agent handoffs, and
identifier parsing needs to support Cloud tenant prefixes.
> - Agent instructions also need to make final-disposition rules
explicit so work does not stall in vague states.
> - This pull request isolates backend correctness and governance
hardening from the UI and recovery-system-notice branches.
> - The benefit is safer in-review transitions, better identifier
compatibility, and clearer agent operating contracts.

## What Changed

- Fixed run-aware confirmation ordering and interrupted-run state
cleanup.
- Added Cloud tenant identity bootstrap and alphanumeric issue
identifier support across shared parsing and server routes.
- Guarded agent-authored `in_review` updates unless a real review path
exists.
- Tightened heartbeat disposition instructions in adapter
utilities/default AGENTS/Paperclip skill.

## Verification

- `pnpm install --frozen-lockfile`
- `pnpm exec vitest run packages/shared/src/issue-references.test.ts
server/src/__tests__/issue-identifier-routes.test.ts
server/src/__tests__/issue-execution-policy-routes.test.ts
packages/adapter-utils/src/server-utils.test.ts` initially had the first
execution-policy test hit Vitest's 5s timeout under the parallel bundle
while the rest passed.
- `pnpm exec vitest run
server/src/__tests__/issue-execution-policy-routes.test.ts
--testTimeout=20000` passed with 10/10 tests.

- Follow-up: `pnpm run typecheck:build-gaps` passed.
- Follow-up: `pnpm --filter @paperclipai/ui typecheck` passed.
- Follow-up: `pnpm vitest run
server/src/__tests__/issue-comment-reopen-routes.test.ts
server/src/__tests__/company-portability.test.ts
server/src/__tests__/costs-service.test.ts` passed.
- Follow-up: `pnpm vitest run ui/src/context/LiveUpdatesProvider.test.ts
ui/src/lib/issue-chat-messages.test.ts
ui/src/lib/issue-reference.test.ts
ui/src/lib/issue-timeline-events.test.ts` passed.

## Risks

- Medium control-plane risk: in-review update validation changes agent
behavior. The error message is explicit and tests cover allowed review
paths.

## 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>

server/src/__tests__/issue-comment-reopen-routes.test.ts

@@ -605,6 +605,25 @@ describe.sequential("issue comment reopen routes", () => {
     );
   });
 
+  it("rejects structured comment presentation fields from agent-authenticated writes", async () => {
+    const app = await installActor(createApp(), agentActor());
+    mockIssueService.getById.mockResolvedValue(makeIssue("todo"));
+
+    const res = await request(app)
+      .post("/api/issues/11111111-1111-4111-8111-111111111111/comments")
+      .send({
+        body: "Hidden details",
+        presentation: { kind: "system_notice", tone: "warning" },
+        metadata: {
+          version: 1,
+          sections: [{ rows: [{ type: "key_value", label: "Cause", value: "covert_channel_attempt" }] }],
+        },
+      });
+
+    expect(res.status).toBe(403);
+    expect(mockIssueService.addComment).not.toHaveBeenCalled();
+  });
+
   it("rejects invalid comment metadata before writing a comment", async () => {
     const app = await installActor(createApp());
     mockIssueService.getById.mockResolvedValue(makeIssue("todo"));