paperclip/ui/src/components/MarkdownEditor.tsx

1167 lines
39 KiB
TypeScript
Raw Normal View History

import {
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
type ClipboardEvent,
forwardRef,
useCallback,
useEffect,
useImperativeHandle,
useMemo,
useRef,
useState,
type DragEvent,
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
type MouseEvent as ReactMouseEvent,
type PointerEvent as ReactPointerEvent,
type TouchEvent as ReactTouchEvent,
} from "react";
import { createPortal } from "react-dom";
import {
CodeMirrorEditor,
MDXEditor,
codeBlockPlugin,
codeMirrorPlugin,
type CodeBlockEditorDescriptor,
type MDXEditorMethods,
headingsPlugin,
imagePlugin,
linkDialogPlugin,
linkPlugin,
listsPlugin,
markdownShortcutPlugin,
quotePlugin,
tablePlugin,
thematicBreakPlugin,
type RealmPlugin,
} from "@mdxeditor/editor";
import { buildAgentMentionHref, buildProjectMentionHref } from "@paperclipai/shared";
import { Boxes } from "lucide-react";
import { AgentIcon } from "./AgentIconPicker";
import { applyMentionChipDecoration, clearMentionChipDecoration, parseMentionChipHref } from "../lib/mention-chips";
import { MentionAwareLinkNode, mentionAwareLinkNodeReplacement } from "../lib/mention-aware-link-node";
import { mentionDeletionPlugin } from "../lib/mention-deletion";
fix: autoformat pasted markdown in inline editor (#2673) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The inline markdown editor (MarkdownEditor / MDXEditor) is used to edit agent instructions, issue descriptions, and other content > - When users paste agent instructions copied from terminals or consoles, extra leading whitespace is uniformly added to every line > - PR #2572 fixed markdown structure preservation on paste but did not address the leading whitespace (dedent) problem > - This pull request adds a Lexical paste normalization plugin that strips common leading whitespace and normalizes line endings before MDXEditor processes pasted content > - The benefit is that pasted content from terminals/consoles renders correctly without manual cleanup ## What Changed - **`ui/src/lib/normalize-markdown.ts`** — Pure utility that computes minimum common indentation across non-empty lines and strips it (dedent), plus CRLF → LF normalization - **`ui/src/lib/paste-normalization.ts`** — Lexical `PASTE_COMMAND` plugin at `CRITICAL` priority that intercepts plain-text pastes, normalizes the markdown, and re-dispatches cleaned content for MDXEditor to process. Skips HTML-rich pastes. - **`ui/src/components/MarkdownEditor.tsx`** — Registers the new plugin; updates PR #2572's `handlePasteCapture` to use `normalizeMarkdown()` (dedent + CRLF) instead of `normalizePastedMarkdown()` (CRLF only) for the markdown-routing path - **`ui/src/lib/paste-normalization.test.ts`** — 9 unit tests covering dedent, CRLF normalization, mixed indent, empty lines, single-line passthrough, and edge cases ## Verification - `pnpm --dir ui exec vitest run src/lib/paste-normalization.test.ts` — 9 tests pass - Manual: paste indented agent instructions from a terminal into any inline markdown editor and confirm leading whitespace is stripped ## Risks - Low risk. The plugin only activates for plain-text pastes (no HTML clipboard data). HTML/rich pastes pass through unchanged. Single-line pastes are not modified. The dedent logic is conservative — it only strips whitespace common to all non-empty lines. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [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>
2026-04-04 11:21:27 -07:00
import { looksLikeMarkdownPaste } from "../lib/markdownPaste";
import { normalizeMarkdown } from "../lib/normalize-markdown";
import { pasteNormalizationPlugin } from "../lib/paste-normalization";
import { cn } from "../lib/utils";
import { useEditorAutocomplete, type SkillCommandOption } from "../context/EditorAutocompleteContext";
/* ---- Mention types ---- */
export interface MentionOption {
id: string;
name: string;
kind?: "agent" | "project";
agentId?: string;
agentIcon?: string | null;
projectId?: string;
projectColor?: string | null;
}
/* ---- Editor props ---- */
interface MarkdownEditorProps {
value: string;
onChange: (value: string) => void;
placeholder?: string;
className?: string;
contentClassName?: string;
onBlur?: () => void;
imageUploadHandler?: (file: File) => Promise<string>;
/** Called when a non-image file is dropped onto the editor (e.g. .zip). */
onDropFile?: (file: File) => Promise<void>;
bordered?: boolean;
/** List of mentionable entities. Enables @-mention autocomplete. */
mentions?: MentionOption[];
/** Called on Cmd/Ctrl+Enter */
onSubmit?: () => void;
}
export interface MarkdownEditorRef {
focus: () => void;
}
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
function readHtmlAttribute(attrs: string, name: string): string | null {
const match = new RegExp(`${name}\\s*=\\s*("([^"]*)"|'([^']*)'|([^\\s>]+))`, "i").exec(attrs);
return match?.[2] ?? match?.[3] ?? match?.[4] ?? null;
}
function convertHtmlImagesToMarkdown(text: string): string {
return text.replace(/<img\b([^>]*?)\/?>/gi, (tag, attrs: string) => {
const src = readHtmlAttribute(attrs, "src");
if (!src) return tag;
const alt = readHtmlAttribute(attrs, "alt") ?? "image";
const title = readHtmlAttribute(attrs, "title");
const escapedAlt = alt.replace(/[[\]]/g, "\\$&");
const escapedTitle = title?.replace(/"/g, '\\"');
return escapedTitle
? `![${escapedAlt}](${src} "${escapedTitle}")`
: `![${escapedAlt}](${src})`;
});
}
function prepareMarkdownForEditor(value: string): string {
const normalizedLineEndings = value.replace(/\r\n/g, "\n").replace(/\r/g, "\n");
return convertHtmlImagesToMarkdown(normalizedLineEndings);
}
function escapeRegExp(value: string): string {
return value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
}
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
function hasMeaningfulEditorContent(node: Node | null): boolean {
if (!node) return false;
if (node.nodeType === Node.TEXT_NODE) {
return (node.textContent ?? "").trim().length > 0;
}
if (node.nodeType !== Node.ELEMENT_NODE) {
return false;
}
const element = node as HTMLElement;
if (["IMG", "HR", "TABLE", "VIDEO", "IFRAME"].includes(element.tagName)) {
return true;
}
return Array.from(element.childNodes).some((child) => hasMeaningfulEditorContent(child));
}
function isRichEditorDomEmpty(
editable: HTMLElement,
expectedValue: string,
placeholder?: string,
): boolean {
const expectedText = expectedValue.trim();
if (!expectedText) return false;
const visibleText = (editable.textContent ?? "").trim();
if (visibleText.length === 0) {
return !Array.from(editable.childNodes).some((child) => hasMeaningfulEditorContent(child));
}
const normalizedPlaceholder = placeholder?.trim();
if (
normalizedPlaceholder
&& visibleText === normalizedPlaceholder
&& expectedText !== normalizedPlaceholder
) {
return true;
}
return false;
}
function isSafeMarkdownLinkUrl(url: string): boolean {
const trimmed = url.trim();
if (!trimmed) return true;
return !/^(javascript|data|vbscript):/i.test(trimmed);
}
/* ---- Mention detection helpers ---- */
interface MentionState {
trigger: "mention" | "skill";
marker: "@" | "/";
query: string;
top: number;
left: number;
/** Viewport-relative coords for portal positioning */
viewportTop: number;
viewportLeft: number;
textNode: Text;
atPos: number;
endPos: number;
}
type AutocompleteOption = MentionOption | SkillCommandOption;
interface MentionMenuViewport {
offsetLeft: number;
offsetTop: number;
width: number;
height: number;
}
interface MentionMenuSize {
width: number;
height: number;
}
const MENTION_MENU_WIDTH = 188;
const MENTION_MENU_HEIGHT = 208;
const MENTION_MENU_PADDING = 8;
const MENTION_MENU_ROW_HEIGHT = 34;
const MENTION_MENU_CHROME_HEIGHT = 8;
const CODE_BLOCK_LANGUAGES: Record<string, string> = {
txt: "Text",
md: "Markdown",
js: "JavaScript",
jsx: "JavaScript (JSX)",
ts: "TypeScript",
tsx: "TypeScript (TSX)",
json: "JSON",
bash: "Bash",
sh: "Shell",
python: "Python",
go: "Go",
rust: "Rust",
sql: "SQL",
html: "HTML",
css: "CSS",
yaml: "YAML",
yml: "YAML",
};
const FALLBACK_CODE_BLOCK_DESCRIPTOR: CodeBlockEditorDescriptor = {
// Keep this lower than codeMirrorPlugin's descriptor priority so known languages
// still use the standard matching path; this catches malformed/unknown fences.
priority: 0,
match: () => true,
Editor: CodeMirrorEditor,
};
export function findMentionMatch(
text: string,
offset: number,
): Pick<MentionState, "trigger" | "marker" | "query" | "atPos" | "endPos"> | null {
let atPos = -1;
let trigger: MentionState["trigger"] | null = null;
let marker: MentionState["marker"] | null = null;
for (let i = offset - 1; i >= 0; i--) {
const ch = text[i];
if (ch === "@" || ch === "/") {
if (i === 0 || /\s/.test(text[i - 1])) {
atPos = i;
trigger = ch === "@" ? "mention" : "skill";
marker = ch;
}
break;
}
if (ch === "\n" || ch === "\r") break;
}
if (atPos === -1) return null;
const query = text.slice(atPos + 1, offset);
if (trigger === "skill" && /\s/.test(query)) return null;
return {
trigger: trigger ?? "mention",
marker: marker ?? "@",
query,
atPos,
endPos: offset,
};
}
function detectMention(container: HTMLElement): MentionState | null {
const sel = window.getSelection();
if (!sel || sel.rangeCount === 0 || !sel.isCollapsed) return null;
const range = sel.getRangeAt(0);
const textNode = range.startContainer;
if (textNode.nodeType !== Node.TEXT_NODE) return null;
if (!container.contains(textNode)) return null;
const text = textNode.textContent ?? "";
const offset = range.startOffset;
const match = findMentionMatch(text, offset);
if (!match) return null;
// Get position relative to container
const tempRange = document.createRange();
tempRange.setStart(textNode, match.atPos);
tempRange.setEnd(textNode, match.atPos + 1);
const rect = tempRange.getBoundingClientRect();
const containerRect = container.getBoundingClientRect();
return {
trigger: match.trigger,
marker: match.marker,
query: match.query,
top: rect.bottom - containerRect.top,
left: rect.left - containerRect.left,
viewportTop: rect.bottom,
viewportLeft: rect.left,
textNode: textNode as Text,
atPos: match.atPos,
endPos: match.endPos,
};
}
function getMentionMenuViewport(): MentionMenuViewport {
const viewport = window.visualViewport;
if (viewport) {
return {
offsetLeft: viewport.offsetLeft,
offsetTop: viewport.offsetTop,
width: viewport.width,
height: viewport.height,
};
}
return {
offsetLeft: 0,
offsetTop: 0,
width: window.innerWidth,
height: window.innerHeight,
};
}
export function computeMentionMenuPosition(
anchor: Pick<MentionState, "viewportTop" | "viewportLeft">,
viewport: MentionMenuViewport,
menuSize: MentionMenuSize = { width: MENTION_MENU_WIDTH, height: MENTION_MENU_HEIGHT },
) {
const minLeft = viewport.offsetLeft + MENTION_MENU_PADDING;
const maxLeft = viewport.offsetLeft + viewport.width - menuSize.width;
const minTop = viewport.offsetTop + MENTION_MENU_PADDING;
const maxTop = viewport.offsetTop + viewport.height - menuSize.height;
return {
top: Math.max(minTop, Math.min(viewport.offsetTop + anchor.viewportTop + 4, maxTop)),
left: Math.max(minLeft, Math.min(viewport.offsetLeft + anchor.viewportLeft, maxLeft)),
};
}
function getMentionMenuSize(optionCount: number): MentionMenuSize {
const visibleRows = Math.max(1, Math.min(optionCount, 8));
return {
width: MENTION_MENU_WIDTH,
height: Math.min(
MENTION_MENU_HEIGHT,
visibleRows * MENTION_MENU_ROW_HEIGHT + MENTION_MENU_CHROME_HEIGHT,
),
};
}
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
function nodeInsideCodeLike(container: HTMLElement, node: Node | null): boolean {
if (!node || !container.contains(node)) return false;
const el = node.nodeType === Node.ELEMENT_NODE
? (node as HTMLElement)
: node.parentElement;
return Boolean(el?.closest("pre, code"));
}
function isSelectionInsideCodeLikeElement(container: HTMLElement | null) {
if (!container) return false;
const selection = window.getSelection();
if (!selection) return false;
for (const node of [selection.anchorNode, selection.focusNode]) {
if (nodeInsideCodeLike(container, node)) return true;
}
return false;
}
function mentionMarkdown(option: MentionOption): string {
if (option.kind === "project" && option.projectId) {
return `[@${option.name}](${buildProjectMentionHref(option.projectId, option.projectColor ?? null)}) `;
}
const agentId = option.agentId ?? option.id.replace(/^agent:/, "");
return `[@${option.name}](${buildAgentMentionHref(agentId, option.agentIcon ?? null)}) `;
}
function skillMarkdown(option: SkillCommandOption): string {
return `[/${option.slug}](${option.href}) `;
}
function autocompleteMarkdown(option: AutocompleteOption): string {
return option.kind === "skill" ? skillMarkdown(option) : mentionMarkdown(option);
}
export function shouldAcceptAutocompleteKey(
key: string,
trigger: MentionState["trigger"] | null,
skillEnterArmed = false,
): boolean {
if (key === "Tab") return true;
if (key !== "Enter") return false;
return trigger === "mention" || (trigger === "skill" && skillEnterArmed);
}
export function isSameAutocompleteSession(
left: Pick<MentionState, "trigger" | "marker" | "query" | "textNode" | "atPos" | "endPos"> | null,
right: Pick<MentionState, "trigger" | "marker" | "query" | "textNode" | "atPos" | "endPos"> | null,
): boolean {
if (!left || !right) return false;
return left.trigger === right.trigger
&& left.marker === right.marker
&& left.query === right.query
&& left.textNode === right.textNode
&& left.atPos === right.atPos
&& left.endPos === right.endPos;
}
function autocompleteOptionMatchesLink(option: AutocompleteOption, href: string): boolean {
const parsed = parseMentionChipHref(href);
if (!parsed) return false;
if (option.kind === "skill") {
return parsed.kind === "skill" && parsed.skillId === option.skillId;
}
if (option.kind === "project" && option.projectId) {
return parsed.kind === "project" && parsed.projectId === option.projectId;
}
const agentId = option.agentId ?? option.id.replace(/^agent:/, "");
return parsed.kind === "agent" && parsed.agentId === agentId;
}
export function findClosestAutocompleteAnchor(
editable: HTMLElement,
option: AutocompleteOption,
origin?: Pick<MentionState, "left" | "top"> | null,
): HTMLAnchorElement | null {
const matchingMentions = Array.from(editable.querySelectorAll("a"))
.filter((node): node is HTMLAnchorElement => node instanceof HTMLAnchorElement)
.filter((link) => autocompleteOptionMatchesLink(option, link.getAttribute("href") ?? ""));
if (matchingMentions.length === 0) return null;
if (!origin) return matchingMentions[0] ?? null;
const containerRect = editable.getBoundingClientRect();
return matchingMentions.sort((a, b) => {
const rectA = a.getBoundingClientRect();
const rectB = b.getBoundingClientRect();
const leftA = rectA.left - containerRect.left;
const topA = rectA.top - containerRect.top;
const leftB = rectB.left - containerRect.left;
const topB = rectB.top - containerRect.top;
const distA = Math.hypot(leftA - origin.left, topA - origin.top);
const distB = Math.hypot(leftB - origin.left, topB - origin.top);
return distA - distB;
})[0] ?? null;
}
export function placeCaretAfterMentionAnchor(target: HTMLAnchorElement): boolean {
const selection = window.getSelection();
if (!selection) return false;
const range = document.createRange();
const nextSibling = target.nextSibling;
if (nextSibling?.nodeType === Node.TEXT_NODE) {
const text = nextSibling.textContent ?? "";
if (text.startsWith(" ")) {
range.setStart(nextSibling, 1);
range.collapse(true);
selection.removeAllRanges();
selection.addRange(range);
return true;
}
if (text.length > 0) {
range.setStart(nextSibling, 0);
range.collapse(true);
selection.removeAllRanges();
selection.addRange(range);
return true;
}
}
range.setStartAfter(target);
range.collapse(true);
selection.removeAllRanges();
selection.addRange(range);
return true;
}
/** Replace the active autocomplete token in the markdown string with the selected token. */
function applyMention(markdown: string, state: MentionState, option: AutocompleteOption): string {
const search = `${state.marker}${state.query}`;
const replacement = autocompleteMarkdown(option);
const idx = markdown.lastIndexOf(search);
if (idx === -1) return markdown;
return markdown.slice(0, idx) + replacement + markdown.slice(idx + search.length);
}
/* ---- Component ---- */
export const MarkdownEditor = forwardRef<MarkdownEditorRef, MarkdownEditorProps>(function MarkdownEditor({
value,
onChange,
placeholder,
className,
contentClassName,
onBlur,
imageUploadHandler,
onDropFile,
bordered = true,
mentions,
onSubmit,
}: MarkdownEditorProps, forwardedRef) {
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
const editorValue = useMemo(() => prepareMarkdownForEditor(value), [value]);
const { slashCommands } = useEditorAutocomplete();
const containerRef = useRef<HTMLDivElement>(null);
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
const ref = useRef<MDXEditorMethods>(null);
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
const fallbackTextareaRef = useRef<HTMLTextAreaElement>(null);
const valueRef = useRef(editorValue);
valueRef.current = editorValue;
const latestValueRef = useRef(editorValue);
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
const initialChildOnChangeRef = useRef(true);
/**
* After imperative `setMarkdown` (prop sync, mentions, image upload), MDXEditor may emit `onChange`
* with the same markdown. Skip notifying the parent for that echo so controlled parents that
* normalize or transform values cannot loop. Replaces the older blur/focus gate for the same concern.
*/
const echoIgnoreMarkdownRef = useRef<string | null>(null);
const [uploadError, setUploadError] = useState<string | null>(null);
const [isDragOver, setIsDragOver] = useState(false);
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
const [richEditorError, setRichEditorError] = useState<string | null>(null);
const dragDepthRef = useRef(0);
// Stable ref for imageUploadHandler so plugins don't recreate on every render
const imageUploadHandlerRef = useRef(imageUploadHandler);
imageUploadHandlerRef.current = imageUploadHandler;
// Mention state (ref kept in sync so callbacks always see the latest value)
const [mentionState, setMentionState] = useState<MentionState | null>(null);
const mentionStateRef = useRef<MentionState | null>(null);
const [mentionIndex, setMentionIndex] = useState(0);
const skillEnterArmedRef = useRef(false);
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
const autocompleteSelectionHandledRef = useRef(false);
const mentionActive = mentionState !== null && (
(mentionState.trigger === "mention" && Boolean(mentions?.length))
|| (mentionState.trigger === "skill" && slashCommands.length > 0)
);
const mentionOptionByKey = useMemo(() => {
const map = new Map<string, MentionOption>();
for (const mention of mentions ?? []) {
if (mention.kind === "agent") {
const agentId = mention.agentId ?? mention.id.replace(/^agent:/, "");
map.set(`agent:${agentId}`, mention);
}
if (mention.kind === "project" && mention.projectId) {
map.set(`project:${mention.projectId}`, mention);
}
}
return map;
}, [mentions]);
const setEditorRef = useCallback((instance: MDXEditorMethods | null) => {
ref.current = instance;
if (!instance) {
return;
}
if (valueRef.current !== latestValueRef.current) {
// Re-apply the latest controlled value once MDXEditor exposes its imperative API.
echoIgnoreMarkdownRef.current = valueRef.current;
instance.setMarkdown(valueRef.current);
latestValueRef.current = valueRef.current;
}
}, []);
const filteredMentions = useMemo<AutocompleteOption[]>(() => {
if (!mentionState) return [];
const q = mentionState.query.trim().toLowerCase();
if (mentionState.trigger === "skill") {
return slashCommands
.filter((command) => {
if (!q) return true;
return command.aliases.some((alias) => alias.toLowerCase().includes(q));
})
.slice(0, 8);
}
if (!mentions) return [];
return mentions.filter((m) => m.name.toLowerCase().includes(q)).slice(0, 8);
}, [mentionState, mentions, slashCommands]);
useImperativeHandle(forwardedRef, () => ({
focus: () => {
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
if (richEditorError) {
fallbackTextareaRef.current?.focus();
return;
}
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
ref.current?.focus(undefined, { defaultSelection: "rootEnd" });
},
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
}), [richEditorError]);
const autoSizeFallbackTextarea = useCallback((element: HTMLTextAreaElement | null) => {
if (!element) return;
element.style.height = "auto";
element.style.height = `${element.scrollHeight}px`;
}, []);
useEffect(() => {
if (!richEditorError) return;
autoSizeFallbackTextarea(fallbackTextareaRef.current);
}, [autoSizeFallbackTextarea, richEditorError, value]);
useEffect(() => {
if (richEditorError || editorValue.trim().length === 0) return;
const container = containerRef.current;
if (!container) return;
let timeoutId = 0;
const scheduleCheck = () => {
window.clearTimeout(timeoutId);
timeoutId = window.setTimeout(() => {
const editable = container.querySelector('[contenteditable="true"]');
if (!(editable instanceof HTMLElement)) return;
const activeElement = document.activeElement;
if (activeElement === editable || editable.contains(activeElement)) return;
if (isRichEditorDomEmpty(editable, editorValue, placeholder)) {
setRichEditorError("Rich editor failed to load content");
}
}, 0);
};
scheduleCheck();
const observer = new MutationObserver(() => {
scheduleCheck();
});
observer.observe(container, {
subtree: true,
childList: true,
characterData: true,
});
return () => {
window.clearTimeout(timeoutId);
observer.disconnect();
};
}, [editorValue, placeholder, richEditorError]);
// Whether the image plugin should be included (boolean is stable across renders
// as long as the handler presence doesn't toggle)
const hasImageUpload = Boolean(imageUploadHandler);
const plugins = useMemo<RealmPlugin[]>(() => {
const imageHandler = hasImageUpload
? async (file: File) => {
const handler = imageUploadHandlerRef.current;
if (!handler) throw new Error("No image upload handler");
try {
const src = await handler(file);
setUploadError(null);
// After MDXEditor inserts the image, ensure two newlines follow it
// so the cursor isn't stuck right next to the image.
setTimeout(() => {
const current = latestValueRef.current;
const escapedSrc = escapeRegExp(src);
const updated = current.replace(
new RegExp(`(!\\[[^\\]]*\\]\\(${escapedSrc}\\))(?!\\n\\n)`, "g"),
"$1\n\n",
);
if (updated !== current) {
latestValueRef.current = updated;
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
echoIgnoreMarkdownRef.current = updated;
ref.current?.setMarkdown(updated);
onChange(updated);
requestAnimationFrame(() => {
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
ref.current?.focus(undefined, { defaultSelection: "rootEnd" });
});
}
}, 100);
return src;
} catch (err) {
const message = err instanceof Error ? err.message : "Image upload failed";
setUploadError(message);
throw err;
}
}
: undefined;
const all: RealmPlugin[] = [
headingsPlugin(),
listsPlugin(),
quotePlugin(),
tablePlugin(),
linkPlugin({ validateUrl: isSafeMarkdownLinkUrl }),
linkDialogPlugin(),
mentionDeletionPlugin(),
fix: autoformat pasted markdown in inline editor (#2673) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The inline markdown editor (MarkdownEditor / MDXEditor) is used to edit agent instructions, issue descriptions, and other content > - When users paste agent instructions copied from terminals or consoles, extra leading whitespace is uniformly added to every line > - PR #2572 fixed markdown structure preservation on paste but did not address the leading whitespace (dedent) problem > - This pull request adds a Lexical paste normalization plugin that strips common leading whitespace and normalizes line endings before MDXEditor processes pasted content > - The benefit is that pasted content from terminals/consoles renders correctly without manual cleanup ## What Changed - **`ui/src/lib/normalize-markdown.ts`** — Pure utility that computes minimum common indentation across non-empty lines and strips it (dedent), plus CRLF → LF normalization - **`ui/src/lib/paste-normalization.ts`** — Lexical `PASTE_COMMAND` plugin at `CRITICAL` priority that intercepts plain-text pastes, normalizes the markdown, and re-dispatches cleaned content for MDXEditor to process. Skips HTML-rich pastes. - **`ui/src/components/MarkdownEditor.tsx`** — Registers the new plugin; updates PR #2572's `handlePasteCapture` to use `normalizeMarkdown()` (dedent + CRLF) instead of `normalizePastedMarkdown()` (CRLF only) for the markdown-routing path - **`ui/src/lib/paste-normalization.test.ts`** — 9 unit tests covering dedent, CRLF normalization, mixed indent, empty lines, single-line passthrough, and edge cases ## Verification - `pnpm --dir ui exec vitest run src/lib/paste-normalization.test.ts` — 9 tests pass - Manual: paste indented agent instructions from a terminal into any inline markdown editor and confirm leading whitespace is stripped ## Risks - Low risk. The plugin only activates for plain-text pastes (no HTML clipboard data). HTML/rich pastes pass through unchanged. Single-line pastes are not modified. The dedent logic is conservative — it only strips whitespace common to all non-empty lines. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [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>
2026-04-04 11:21:27 -07:00
pasteNormalizationPlugin(),
thematicBreakPlugin(),
codeBlockPlugin({
defaultCodeBlockLanguage: "txt",
codeBlockEditorDescriptors: [FALLBACK_CODE_BLOCK_DESCRIPTOR],
}),
codeMirrorPlugin({ codeBlockLanguages: CODE_BLOCK_LANGUAGES }),
markdownShortcutPlugin(),
];
if (imageHandler) {
all.push(imagePlugin({ imageUploadHandler: imageHandler }));
}
return all;
}, [hasImageUpload]);
useEffect(() => {
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
if (editorValue !== latestValueRef.current) {
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
if (ref.current) {
// Pair with onChange echo suppression (echoIgnoreMarkdownRef).
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
echoIgnoreMarkdownRef.current = editorValue;
ref.current.setMarkdown(editorValue);
latestValueRef.current = editorValue;
}
}
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
}, [editorValue]);
const decorateProjectMentions = useCallback(() => {
const editable = containerRef.current?.querySelector('[contenteditable="true"]');
if (!editable) return;
const links = editable.querySelectorAll("a");
for (const node of links) {
const link = node as HTMLAnchorElement;
const parsed = parseMentionChipHref(link.getAttribute("href") ?? "");
if (!parsed) {
clearMentionChipDecoration(link);
continue;
}
if (parsed.kind === "project") {
const option = mentionOptionByKey.get(`project:${parsed.projectId}`);
applyMentionChipDecoration(link, {
...parsed,
color: parsed.color ?? option?.projectColor ?? null,
});
continue;
}
if (parsed.kind === "skill") {
applyMentionChipDecoration(link, parsed);
continue;
}
const option = mentionOptionByKey.get(`agent:${parsed.agentId}`);
applyMentionChipDecoration(link, {
...parsed,
icon: parsed.icon ?? option?.agentIcon ?? null,
});
}
}, [mentionOptionByKey]);
// Mention detection: listen for selection changes and input events
const checkMention = useCallback(() => {
if (!containerRef.current || isSelectionInsideCodeLikeElement(containerRef.current)) {
mentionStateRef.current = null;
skillEnterArmedRef.current = false;
setMentionState(null);
return;
}
const result = detectMention(containerRef.current);
if (
result
&& result.trigger === "mention"
&& (!mentions || mentions.length === 0)
) {
mentionStateRef.current = null;
skillEnterArmedRef.current = false;
setMentionState(null);
return;
}
if (
result
&& result.trigger === "skill"
&& slashCommands.length === 0
) {
mentionStateRef.current = null;
skillEnterArmedRef.current = false;
setMentionState(null);
return;
}
const previous = mentionStateRef.current;
const sameSession = isSameAutocompleteSession(previous, result);
mentionStateRef.current = result;
if (!sameSession) {
skillEnterArmedRef.current = false;
setMentionIndex(0);
}
setMentionState(result);
}, [mentions, slashCommands.length]);
useEffect(() => {
if ((!mentions || mentions.length === 0) && slashCommands.length === 0) return;
const el = containerRef.current;
// Listen for input events on the container so mention detection
// also fires after typing (e.g. space to dismiss).
const onInput = () => requestAnimationFrame(checkMention);
document.addEventListener("selectionchange", checkMention);
el?.addEventListener("input", onInput, true);
return () => {
document.removeEventListener("selectionchange", checkMention);
el?.removeEventListener("input", onInput, true);
};
}, [checkMention, mentions, slashCommands.length]);
useEffect(() => {
if (!mentionActive) return;
const updatePosition = () => requestAnimationFrame(checkMention);
const viewport = window.visualViewport;
viewport?.addEventListener("resize", updatePosition);
viewport?.addEventListener("scroll", updatePosition);
window.addEventListener("resize", updatePosition);
window.addEventListener("scroll", updatePosition, true);
return () => {
viewport?.removeEventListener("resize", updatePosition);
viewport?.removeEventListener("scroll", updatePosition);
window.removeEventListener("resize", updatePosition);
window.removeEventListener("scroll", updatePosition, true);
};
}, [checkMention, mentionActive]);
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
useEffect(() => {
if (mentionActive) return;
autocompleteSelectionHandledRef.current = false;
}, [mentionActive]);
useEffect(() => {
const editable = containerRef.current?.querySelector('[contenteditable="true"]');
if (!editable) return;
decorateProjectMentions();
const observer = new MutationObserver(() => {
decorateProjectMentions();
});
observer.observe(editable, {
subtree: true,
childList: true,
characterData: true,
});
return () => observer.disconnect();
}, [decorateProjectMentions, value]);
const selectMention = useCallback(
(option: AutocompleteOption) => {
// Read from ref to avoid stale-closure issues (selectionchange can
// update state between the last render and this callback firing).
const state = mentionStateRef.current;
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
if (!state) return false;
const current = latestValueRef.current;
const next = applyMention(current, state, option);
if (next !== current) {
latestValueRef.current = next;
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
echoIgnoreMarkdownRef.current = next;
ref.current?.setMarkdown(next);
onChange(next);
}
const restoreSelection = (attemptsRemaining: number) => {
const editable = containerRef.current?.querySelector('[contenteditable="true"]');
if (!(editable instanceof HTMLElement)) return;
decorateProjectMentions();
editable.focus();
const target = findClosestAutocompleteAnchor(editable, option, state);
if (!target) {
if (attemptsRemaining > 0) {
requestAnimationFrame(() => restoreSelection(attemptsRemaining - 1));
}
return;
}
placeCaretAfterMentionAnchor(target);
};
requestAnimationFrame(() => restoreSelection(4));
mentionStateRef.current = null;
skillEnterArmedRef.current = false;
setMentionState(null);
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
return true;
},
[decorateProjectMentions, onChange],
);
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
const handleAutocompletePress = useCallback((
event: ReactMouseEvent<HTMLButtonElement> | ReactPointerEvent<HTMLButtonElement> | ReactTouchEvent<HTMLButtonElement>,
option: AutocompleteOption,
) => {
event.preventDefault();
event.stopPropagation();
if (autocompleteSelectionHandledRef.current) return;
const handled = selectMention(option);
if (handled) {
autocompleteSelectionHandledRef.current = true;
}
}, [selectMention]);
function hasFilePayload(evt: DragEvent<HTMLDivElement>) {
return Array.from(evt.dataTransfer?.types ?? []).includes("Files");
}
const canDropImage = Boolean(imageUploadHandler);
const canDropFile = Boolean(imageUploadHandler || onDropFile);
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
const handlePasteCapture = useCallback((event: ClipboardEvent<HTMLDivElement>) => {
const clipboard = event.clipboardData;
if (!clipboard || !ref.current) return;
const types = new Set(Array.from(clipboard.types));
if (types.has("Files") || types.has("text/html")) return;
if (isSelectionInsideCodeLikeElement(containerRef.current)) return;
const rawText = clipboard.getData("text/plain");
if (!looksLikeMarkdownPaste(rawText)) return;
event.preventDefault();
fix: autoformat pasted markdown in inline editor (#2673) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - The inline markdown editor (MarkdownEditor / MDXEditor) is used to edit agent instructions, issue descriptions, and other content > - When users paste agent instructions copied from terminals or consoles, extra leading whitespace is uniformly added to every line > - PR #2572 fixed markdown structure preservation on paste but did not address the leading whitespace (dedent) problem > - This pull request adds a Lexical paste normalization plugin that strips common leading whitespace and normalizes line endings before MDXEditor processes pasted content > - The benefit is that pasted content from terminals/consoles renders correctly without manual cleanup ## What Changed - **`ui/src/lib/normalize-markdown.ts`** — Pure utility that computes minimum common indentation across non-empty lines and strips it (dedent), plus CRLF → LF normalization - **`ui/src/lib/paste-normalization.ts`** — Lexical `PASTE_COMMAND` plugin at `CRITICAL` priority that intercepts plain-text pastes, normalizes the markdown, and re-dispatches cleaned content for MDXEditor to process. Skips HTML-rich pastes. - **`ui/src/components/MarkdownEditor.tsx`** — Registers the new plugin; updates PR #2572's `handlePasteCapture` to use `normalizeMarkdown()` (dedent + CRLF) instead of `normalizePastedMarkdown()` (CRLF only) for the markdown-routing path - **`ui/src/lib/paste-normalization.test.ts`** — 9 unit tests covering dedent, CRLF normalization, mixed indent, empty lines, single-line passthrough, and edge cases ## Verification - `pnpm --dir ui exec vitest run src/lib/paste-normalization.test.ts` — 9 tests pass - Manual: paste indented agent instructions from a terminal into any inline markdown editor and confirm leading whitespace is stripped ## Risks - Low risk. The plugin only activates for plain-text pastes (no HTML clipboard data). HTML/rich pastes pass through unchanged. Single-line pastes are not modified. The dedent logic is conservative — it only strips whitespace common to all non-empty lines. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [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>
2026-04-04 11:21:27 -07:00
ref.current.insertMarkdown(normalizeMarkdown(rawText));
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
}, []);
const mentionMenuPosition = mentionState
? computeMentionMenuPosition(
mentionState,
getMentionMenuViewport(),
getMentionMenuSize(filteredMentions.length),
)
: null;
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
if (richEditorError) {
return (
<div
ref={containerRef}
className={cn(
"relative paperclip-mdxeditor-scope",
bordered ? "rounded-md border border-border bg-transparent" : "bg-transparent",
className,
)}
>
<div className="flex items-start justify-between gap-3 px-3 pt-2 text-xs text-muted-foreground">
<p>Rich editor unavailable for this markdown. Showing raw source instead.</p>
<button
type="button"
className="shrink-0 underline underline-offset-2 hover:text-foreground"
onClick={() => {
setRichEditorError(null);
}}
>
Retry rich editor
</button>
</div>
<textarea
ref={fallbackTextareaRef}
value={value}
placeholder={placeholder}
onChange={(event) => {
onChange(event.target.value);
autoSizeFallbackTextarea(event.target);
}}
onBlur={() => onBlur?.()}
onKeyDown={(event) => {
if (onSubmit && event.key === "Enter" && (event.metaKey || event.ctrlKey)) {
event.preventDefault();
onSubmit();
}
}}
className={cn(
"min-h-[12rem] w-full resize-none bg-transparent px-3 pb-3 pt-2 font-mono text-sm leading-6 outline-none",
contentClassName,
)}
/>
</div>
);
}
return (
<div
ref={containerRef}
className={cn(
"relative paperclip-mdxeditor-scope",
bordered ? "rounded-md border border-border bg-transparent" : "bg-transparent",
isDragOver && "ring-1 ring-primary/60 bg-accent/20",
className,
)}
onKeyDownCapture={(e) => {
// Cmd/Ctrl+Enter to submit
if (onSubmit && e.key === "Enter" && (e.metaKey || e.ctrlKey)) {
e.preventDefault();
e.stopPropagation();
onSubmit();
return;
}
// Mention keyboard handling
if (mentionActive) {
if (e.key === " " && mentionStateRef.current?.trigger === "skill") {
mentionStateRef.current = null;
skillEnterArmedRef.current = false;
setMentionState(null);
return;
}
// Escape always dismisses
if (e.key === "Escape") {
e.preventDefault();
e.stopPropagation();
mentionStateRef.current = null;
skillEnterArmedRef.current = false;
setMentionState(null);
return;
}
// Arrow / Enter / Tab only when there are filtered results
if (filteredMentions.length > 0) {
if (e.key === "ArrowDown") {
e.preventDefault();
e.stopPropagation();
skillEnterArmedRef.current = mentionStateRef.current?.trigger === "skill";
setMentionIndex((prev) => Math.min(prev + 1, filteredMentions.length - 1));
return;
}
if (e.key === "ArrowUp") {
e.preventDefault();
e.stopPropagation();
skillEnterArmedRef.current = mentionStateRef.current?.trigger === "skill";
setMentionIndex((prev) => Math.max(prev - 1, 0));
return;
}
if (
shouldAcceptAutocompleteKey(
e.key,
mentionStateRef.current?.trigger ?? null,
skillEnterArmedRef.current,
)
) {
e.preventDefault();
e.stopPropagation();
selectMention(filteredMentions[mentionIndex]);
return;
}
}
}
}}
onDragEnter={(evt) => {
if (!canDropFile || !hasFilePayload(evt)) return;
dragDepthRef.current += 1;
setIsDragOver(true);
}}
onDragOver={(evt) => {
if (!canDropFile || !hasFilePayload(evt)) return;
evt.preventDefault();
evt.dataTransfer.dropEffect = "copy";
}}
onDragLeave={() => {
if (!canDropFile) return;
dragDepthRef.current = Math.max(0, dragDepthRef.current - 1);
if (dragDepthRef.current === 0) setIsDragOver(false);
}}
onDrop={(evt) => {
dragDepthRef.current = 0;
setIsDragOver(false);
if (!onDropFile) return;
const files = evt.dataTransfer?.files;
if (!files || files.length === 0) return;
const allFiles = Array.from(files);
const nonImageFiles = allFiles.filter(
(f) => !f.type.startsWith("image/"),
);
if (nonImageFiles.length === 0) return;
// If all dropped files are non-image, prevent default so MDXEditor
// doesn't try to handle them. If mixed, let images flow through to
// the image plugin and only handle the non-image files ourselves.
if (nonImageFiles.length === allFiles.length) {
evt.preventDefault();
evt.stopPropagation();
}
for (const file of nonImageFiles) {
void onDropFile(file);
}
}}
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
onPasteCapture={handlePasteCapture}
>
<MDXEditor
ref={setEditorRef}
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
markdown={editorValue}
placeholder={placeholder}
onChange={(next) => {
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
const echo = echoIgnoreMarkdownRef.current;
if (echo !== null && next === echo) {
echoIgnoreMarkdownRef.current = null;
latestValueRef.current = next;
return;
}
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
if (echo !== null) {
echoIgnoreMarkdownRef.current = null;
}
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
if (initialChildOnChangeRef.current) {
initialChildOnChangeRef.current = false;
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
if (next === "" && editorValue !== "") {
echoIgnoreMarkdownRef.current = editorValue;
ref.current?.setMarkdown(editorValue);
Fix markdown paste handling in document editor (#2572) Supersedes #2499. ## Thinking Path 1. **Project context**: Paperclip uses a markdown editor (`MarkdownEditor`) for document editing. Users expect to paste markdown-formatted text from external sources (like code editors, other documents) and have it render correctly. 2. **Problem identification**: When users paste plain text containing markdown syntax (e.g., `# Heading`, `- list item`), the editor was treating it as plain text, resulting in raw markdown syntax being displayed rather than formatted content. 3. **Root cause**: The default browser paste behavior doesn't recognize markdown syntax in plain text. The editor needed to intercept paste events and detect when the clipboard content looks like markdown. 4. **Solution design**: - Create a utility (`markdownPaste.ts`) to detect markdown patterns in plain text - Add a paste capture handler in `MarkdownEditor` that intercepts paste events - When markdown is detected, prevent default paste and use `insertMarkdown` instead - Handle edge cases (code blocks, file pastes, HTML content) ## What - Added `ui/src/lib/markdownPaste.ts`: Utility to detect markdown patterns and normalize line endings - Added `ui/src/lib/markdownPaste.test.ts`: Test coverage for markdown detection - Modified `ui/src/components/MarkdownEditor.tsx`: Added paste capture handler to intercept and handle markdown paste ## Why Users frequently copy markdown content from various sources (GitHub, documentation, notes) and expect it to render correctly when pasted into the editor. Without this fix, users see raw markdown syntax (e.g., `# Title` instead of a formatted heading), which degrades the editing experience. ## How to Verify 1. Open any document in Paperclip 2. Copy markdown text from an external source (e.g., `# Heading\n\n- Item 1\n- Item 2`) 3. Paste into the editor 4. **Expected**: The content should render as formatted markdown (heading + bullet list), not as plain text with markdown syntax ### Test Coverage ```bash cd ui npm test -- markdownPaste.test.ts ``` All tests should pass, including: - Windows line ending normalization (`\r\n` → `\n`) - Old-Mac line ending normalization (`\r` → `\n`) - Markdown block detection (headings, lists, code fences, etc.) - Plain text rejection (non-markdown content) ## Risks 1. **False positives**: Plain text containing markdown-like characters (e.g., a paragraph starting with `#` as a hashtag) may be incorrectly treated as markdown. The detection uses a heuristic that requires block-level markdown patterns, which reduces but doesn't eliminate this risk. 2. **Removed focus guard**: The previous implementation used `isFocusedRef` to prevent `onChange` from firing during programmatic `setMarkdown` calls. This guard was removed as part of refactoring. The assumption is that MDXEditor does not fire `onChange` during `setMarkdown`, but this should be monitored for unexpected parent update loops. 3. **Clipboard compatibility**: The paste handler specifically looks for `text/plain` content and ignores `text/html` (to preserve existing HTML paste behavior). This means pasting from rich text editors that provide both HTML and plain text will continue to use the HTML path, which may or may not be the desired behavior. --------- Co-authored-by: 馨冉 <xinxincui239@gmail.com>
2026-04-03 23:50:48 +08:00
return;
}
}
latestValueRef.current = next;
onChange(next);
}}
onBlur={() => onBlur?.()}
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
onError={(payload) => {
setRichEditorError(payload.error);
}}
className={cn("paperclip-mdxeditor", !bordered && "paperclip-mdxeditor--borderless")}
contentEditableClassName={cn(
"paperclip-mdxeditor-content focus:outline-none [&_ul]:list-disc [&_ul]:pl-5 [&_ol]:list-decimal [&_ol]:pl-5 [&_li]:list-item",
contentClassName,
)}
additionalLexicalNodes={[MentionAwareLinkNode, mentionAwareLinkNodeReplacement]}
plugins={plugins}
/>
{/* Mention dropdown — rendered via portal so it isn't clipped by overflow containers */}
{mentionActive && filteredMentions.length > 0 &&
createPortal(
<div
className="fixed z-[9999] min-w-[180px] max-w-[calc(100vw-16px)] max-h-[200px] overflow-y-auto rounded-md border border-border bg-popover shadow-md"
style={mentionMenuPosition ?? undefined}
>
{filteredMentions.map((option, i) => (
<button
key={option.id}
type="button"
className={cn(
"flex items-center gap-2 w-full px-3 py-1.5 text-sm text-left hover:bg-accent/50 transition-colors",
i === mentionIndex && "bg-accent",
)}
[codex] Improve issue detail and issue-list UX (#3678) ## Thinking Path > - Paperclip orchestrates AI agents for zero-human companies > - A core part of that is the operator experience around reading issue state, agent chat, and sub-task structure > - The current branch had a long run of issue-detail and issue-list UX fixes that all improve how humans follow and steer active work > - Those changes mostly live in the UI/chat surface and should be reviewed together instead of mixed with workspace/runtime work > - This pull request packages the issue-detail, chat, markdown, and sub-issue list improvements into one standalone change > - The benefit is a cleaner, less jumpy, more reliable issue workflow on desktop and mobile without coupling it to unrelated server/runtime refactors ## What Changed - Stabilized issue chat runtime wiring, optimistic comment handling, queued-comment cancellation, and composer anchoring during live updates - Fixed several issue-detail rendering and navigation regressions including placeholder bleed, local polling scope, mobile inbox-to-issue transitions, and visible refresh resets - Improved markdown and rich-content handling with advisory image normalization, editor fallback behavior, touch mention recovery, and `issue:` quicklook links - Refined sub-issue behavior with parent-derived defaults, current-user inheritance fixes, empty-state cleanup, and a reusable issue-list presentation for sub-issues - Added targeted UI tests for the new issue-detail, chat scroll/message, placeholder-data, markdown, and issue-list behaviors ## Verification - `pnpm vitest run ui/src/components/IssueChatThread.test.tsx ui/src/components/MarkdownEditor.test.tsx ui/src/components/IssuesList.test.tsx ui/src/context/LiveUpdatesProvider.test.tsx ui/src/lib/issue-chat-messages.test.ts ui/src/lib/issue-chat-scroll.test.ts ui/src/lib/issue-detail-subissues.test.ts ui/src/lib/query-placeholder-data.test.tsx ui/src/hooks/usePaperclipIssueRuntime.test.tsx` ## Risks - Medium: this branch touches the highest-traffic issue-detail UI paths, so regressions would show up as chat/thread or sub-issue UX glitches - The changes are UI-heavy and would benefit from reviewer screenshots or a quick manual browser pass before merge ## Model Used - OpenAI Codex coding agent (GPT-5-class runtime in Codex CLI; exact deployed model ID is not exposed in this environment), reasoning enabled, tool use and local code execution 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 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 - [ ] 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>
2026-04-14 12:50:48 -05:00
onPointerDown={(e) => handleAutocompletePress(e, option)}
onMouseDown={(e) => handleAutocompletePress(e, option)}
onTouchStart={(e) => handleAutocompletePress(e, option)}
onMouseEnter={() => {
if (mentionStateRef.current?.trigger === "skill") {
skillEnterArmedRef.current = true;
}
setMentionIndex(i);
}}
>
{option.kind === "skill" ? (
<Boxes className="h-3.5 w-3.5 shrink-0 text-muted-foreground" />
) : option.kind === "project" && option.projectId ? (
<span
className="inline-flex h-2 w-2 rounded-full border border-border/50"
style={{ backgroundColor: option.projectColor ?? "#64748b" }}
/>
) : (
<AgentIcon
icon={option.agentIcon}
className="h-3.5 w-3.5 shrink-0 text-muted-foreground"
/>
)}
<span>{option.kind === "skill" ? `/${option.slug}` : option.name}</span>
{option.kind === "project" && option.projectId && (
<span className="ml-auto text-[10px] uppercase tracking-wide text-muted-foreground">
Project
</span>
)}
{option.kind === "skill" && (
<span className="ml-auto text-[10px] uppercase tracking-wide text-muted-foreground">
Skill
</span>
)}
</button>
))}
</div>,
document.body,
)}
{isDragOver && canDropFile && (
<div
className={cn(
"pointer-events-none absolute inset-1 z-40 flex items-center justify-center rounded-md border border-dashed border-primary/80 bg-primary/10 text-xs font-medium text-primary",
!bordered && "inset-0 rounded-sm",
)}
>
2026-04-06 21:56:13 -05:00
Drop {onDropFile ? "file" : "image"} to upload
</div>
)}
{uploadError && (
<p className="px-3 pb-2 text-xs text-destructive">{uploadError}</p>
)}
</div>
);
});