[qdaxb]

Summary

• restore a conversation transcript when the loaded task key matches but the pane is displaying another task
• add a regression test for switching A → pending B → A before B finishes loading
• extend the real desktop short-conversation E2E with a running second conversation and rapid switching
• document the transcript identity rule in Chinese and English

Root cause

loadedRuntimeTranscriptKeyRef tracked only the last successfully loaded task. During a rapid switch, task B could replace the displayed messages while its transcript was still pending. Switching back to task A then returned early because A was still the last loaded key, even though the pane was displaying B.

The pane now skips transcript restoration only when both the loaded key and the displayed transcript identity match the target task.

Verification

• pnpm --filter wework exec vitest run src/components/layout/useWorkbenchPaneSession.test.ts src/features/workbench/WorkbenchProvider.test.tsx — 142 passed
• focused race regression after rebasing onto latest origin/main — passed
• Prettier and ESLint for all changed Wework files — passed
• real Tauri short-conversation E2E — passed
  • evidence: wework/test-results/desktop-e2e/2026-07-30T08-06-18-113Z-11406
• pre-push ESLint, TypeScript, and unit-test suite — passed

Summary by CodeRabbit

• Bug Fixes

  • Improved conversation switching so cached history restores reliably when returning before another transcript finishes loading.
  • Prevented delayed transcript responses from overwriting the currently displayed conversation.
  • Ensured transcript restoration only skips reload when the displayed conversation matches the loaded transcript identity, keeping historical turns visible.

• Tests

  • Added component race-condition coverage for rapid runtime pane switching and transcript restoration.
  • Extended desktop E2E flows to verify prompt/completion restoration and that background conversation transcripts do not leak into the restored view.

Additional CI fix

• classify pull request changes from the merge base in lint, test, platform E2E, and Wework E2E workflows
• prevent files added to main after a branch diverges from incorrectly enabling backend, CLI, frontend, or platform E2E jobs
• add a workflow-level regression assertion for three-dot diff usage

Additional verification

• bash .github/scripts/test-classify-ci-changes.sh — passed
• actionlint -shellcheck= for all changed workflows — passed
• PR fix(wework): restore history after rapid conversation switching #2333 base/head reproduction now detects 6 actual PR files instead of 142 tree differences

[coderabbitai]

📝 Walkthrough

Walkthrough

Runtime transcript restoration now requires both transcript key and displayed identity matches. Component and desktop coverage exercise rapid switching and late-response isolation. CI path classification now uses merge-base diff ranges, with validation ensuring all relevant workflows use the updated syntax.

Changes

Transcript restoration and desktop flow validation

Layer / File(s)	Summary
Identity-aware transcript loading wework/src/components/layout/useWorkbenchPaneSession.ts, docs/en/..., docs/zh/...	Transcript reloads are skipped only when the loaded key and displayed identity both match the target; corresponding restoration contracts are documented.
Component race validation wework/src/features/workbench/WorkbenchProvider.test.tsx	A runtime switching test verifies cached history restoration and isolation from a late transcript response.
Desktop switching and flow validation wework/src/e2e/automation.ts, wework/e2e/desktop/task-flow.e2e.mjs	Adds macrotask-ordered clicking and validates historical message restoration, task isolation, and task-row reuse during concurrent conversations.

CI changed-path classification

Layer / File(s)	Summary
Merge-base workflow classification .github/workflows/{lint,test,e2e-tests,wework-e2e}.yml	Changed-path detection uses three-dot merge-base ranges for downstream job classification.
Classification syntax validation .github/scripts/test-classify-ci-changes.sh	The classifier test verifies all four workflows contain the required merge-base diff command.

Estimated code review effort: 3 (Moderate) | ~25 minutes

Possibly related PRs

• wecode-ai/Wegent#2165: Updates related keyed runtime conversation cache and restoration behavior.
• wecode-ai/Wegent#2248: Modifies the same CI changed-path classification machinery.
• wecode-ai/Wegent#2281: Changes related transcript state handling during background task switching.

Suggested reviewers: micro66, ficohu

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly matches the main change: fixing transcript history restoration during rapid conversation switching in Wework.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/wework-conversation-history-switch

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@wework/e2e/desktop/task-flow.e2e.mjs`:
- Around line 1597-1618: Move or repeat the active-pane snapshot assertions
after releaseConcurrentMemoryResponses() and the running-task settlement in the
rapid-switch flow. Verify the active A conversation still contains both expected
user and assistant history messages and excludes
CONVERSATION_SWITCH_RACE_PROMPT, using the existing snapshot and assertion
helpers around restoredAfterRace.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 68b54652-ef18-48bf-b408-fb985c05452a

📥 Commits

Reviewing files that changed from the base of the PR and between d43c5c9 and 5a1030f.

📒 Files selected for processing (6)

• docs/en/wework/developer-guide/wework-chat-state-sources.md
• docs/zh/wework/developer-guide/wework-chat-state-sources.md
• wework/e2e/desktop/task-flow.e2e.mjs
• wework/src/components/layout/useWorkbenchPaneSession.ts
• wework/src/e2e/automation.ts
• wework/src/features/workbench/WorkbenchProvider.test.tsx

🚧 Files skipped from review as they are similar to previous changes (2)

• docs/zh/wework/developer-guide/wework-chat-state-sources.md
• docs/en/wework/developer-guide/wework-chat-state-sources.md

[coderabbitai]

🧹 Nitpick comments (1)

wework/e2e/desktop/task-flow.e2e.mjs (1)

1562-1566: 📐 Maintainability & Code Quality | 🔵 Trivial | 💤 Low value

Repeated task-row snapshot boilerplate.

taskRowsBeforeRace duplicates the same "snapshot body → filter runtime-local-task-row- ids" pattern already used for taskRowsBeforeConversation (line 1442) and elsewhere in this file (taskRowsBeforeInitialTask, projectRowsBeforeComposerCreate). Extracting a small helper (e.g. captureExistingTaskRowIds(control)) would remove this recurring duplication.

As per coding guidelines: "Before adding code, search for and reuse existing components, services, utilities, and patterns; extract shared logic instead of duplicating it."

♻️ Suggested helper

+async function captureExistingTaskRowIds(control) {
+  return new Set(
+    JSON.parse(await control.command('snapshot', 'body')).testIds.filter(testId =>
+      testId.startsWith('runtime-local-task-row-')
+    )
+  )
+}

-  const taskRowsBeforeRace = new Set(
-    JSON.parse(await control.command('snapshot', 'body')).testIds.filter(testId =>
-      testId.startsWith('runtime-local-task-row-')
-    )
-  )
+  const taskRowsBeforeRace = await captureExistingTaskRowIds(control)

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@wework/e2e/desktop/task-flow.e2e.mjs` around lines 1562 - 1566, Extract the
repeated snapshot-and-filter logic into a shared helper such as
captureExistingTaskRowIds(control), then replace taskRowsBeforeRace and the
equivalent task-row collection sites including taskRowsBeforeConversation and
taskRowsBeforeInitialTask with calls to that helper. Preserve the existing
snapshot command and runtime-local-task-row- ID filtering behavior.

Source: Coding guidelines

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In `@wework/e2e/desktop/task-flow.e2e.mjs`:
- Around line 1562-1566: Extract the repeated snapshot-and-filter logic into a
shared helper such as captureExistingTaskRowIds(control), then replace
taskRowsBeforeRace and the equivalent task-row collection sites including
taskRowsBeforeConversation and taskRowsBeforeInitialTask with calls to that
helper. Preserve the existing snapshot command and runtime-local-task-row- ID
filtering behavior.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 186a9547-7bdf-4686-8200-92da388a9a44

📥 Commits

Reviewing files that changed from the base of the PR and between 5a1030f and 38d2727.

📒 Files selected for processing (11)

• .github/scripts/test-classify-ci-changes.sh
• .github/workflows/e2e-tests.yml
• .github/workflows/lint.yml
• .github/workflows/test.yml
• .github/workflows/wework-e2e.yml
• docs/en/wework/developer-guide/wework-chat-state-sources.md
• docs/zh/wework/developer-guide/wework-chat-state-sources.md
• wework/e2e/desktop/task-flow.e2e.mjs
• wework/src/components/layout/useWorkbenchPaneSession.ts
• wework/src/e2e/automation.ts
• wework/src/features/workbench/WorkbenchProvider.test.tsx

🚧 Files skipped from review as they are similar to previous changes (1)

• docs/en/wework/developer-guide/wework-chat-state-sources.md