fix(feishu): reply inside thread for P2P direct messages by LiaoyuanNing · Pull Request #38808 · openclaw/openclaw

LiaoyuanNing · 2026-03-07T11:24:02Z

Summary

Fix bot replying outside thread in P2P (direct message) chats when user sends a message inside a Feishu thread
Distinguish between thread messages (thread_id present) and plain quote replies (root_id only) per the Feishu API contract
Group chat behavior is completely unchanged — the new condition requires isDirect=true

Changes

Four targeted changes in extensions/feishu/src/bot.ts, all gated by directThreadMessage = isDirect && Boolean(ctx.threadId):

Parameter	Before (P2P)	After (P2P thread)	After (P2P non-thread)
`replyInThread`	`false`	`true`	`false` (unchanged)
`replyTargetMessageId`	`ctx.messageId`	`ctx.rootId`	`ctx.messageId` (unchanged)
`threadReply`	`false`	`true`	`false` (unchanged)
`skipReplyToInMessages`	`true`	`false`	`true` (unchanged)

Test plan

Added test: P2P thread message (thread_id present) → replies to root with replyInThread=true
Added test: P2P plain reply (root_id only, no thread_id) → replies to triggering message with replyInThread=false
All 54 existing feishu bot tests pass
Manual verification: send message in P2P thread, confirm bot replies inside thread

greptile-apps · 2026-03-07T11:26:26Z

Greptile Summary

This PR fixes a Feishu bot regression where replies to P2P (direct message) thread messages were posted outside the thread container. The fix introduces a directThreadMessage flag (isDirect && Boolean(ctx.threadId)) to distinguish Feishu thread messages from plain quote replies in P2P chats, and uses it to correctly set replyInThread, threadReply, replyTargetMessageId, and skipReplyToInMessages — only for the thread case.

The directThreadMessage guard is correctly scoped to isDirect, so group chat behavior is completely unaffected. Both broadcast and non-broadcast dispatcher call sites are updated consistently. The fallback ctx.rootId ?? ctx.messageId gracefully handles any edge case where rootId is absent despite threadId being present. Two new tests cover the P2P-thread and P2P-plain-reply paths with precise assertions; no existing tests were modified.

Confidence Score: 5/5

This PR is safe to merge — the change is minimal, well-guarded by an isDirect condition, and covered by two new targeted tests.
The fix is narrow (four lines of logic under a single boolean flag), does not affect group chats at all, preserves the existing P2P non-thread behavior exactly, and is backed by two new tests that assert the correct parameters for each case. No risky patterns such as removed validations, new async paths, or external API contract changes were introduced.
No files require special attention.

_{Last reviewed commit: 1bd3555}

LiaoyuanNing · 2026-03-09T03:27:03Z

Hi @steipete @onutc, this is a small fix for Feishu P2P thread replies. Greptile bot confirmed it's safe to merge. Could you take a look when you have a chance? Thanks!

openclaw-barnacle · 2026-04-22T04:40:27Z

This pull request has been automatically marked as stale due to inactivity.
Please add updates or it will be closed.

clawsweeper · 2026-04-26T01:47:37Z

Codex review: keeping this open for maintainer follow-up; there is still a little grit to resolve.

Keep this PR open. Current main still routes Feishu P2P direct-thread messages through the non-thread DM path, while the paired open bug #38806 remains unresolved and this PR is the active narrow implementation candidate. The provided PR diff is scoped to Feishu bot routing, tests, and changelog, with no visible security-sensitive or supply-chain changes.

Best possible solution:

Review and land this narrow Feishu plugin fix, or apply the same behavior on main: detect direct messages with thread_id, reply to root_id with replyInThread/threadReply enabled, keep root_id-only P2P quote replies on the non-thread path, preserve existing group behavior, and keep focused regression tests. After merge, close #38806 as implemented.

What I checked:

current-main-p2p-reply-in-thread-missing: Current main still sets replyInThread to false for every non-group Feishu message, so a P2P event with thread_id cannot request Feishu reply_in_thread through this path. (extensions/feishu/src/bot.ts:704, ef58307f843e)
current-main-reply-target-and-thread-reply-group-gated: The reply target chooses a root only for topic/configured group sessions, and threadReply is still false for non-groups, leaving P2P thread messages outside the thread-aware branch. (extensions/feishu/src/bot.ts:1214, ef58307f843e)
current-main-dispatch-skips-p2p-reply-metadata: Both dispatcher call sites still pass skipReplyToInMessages: !isGroup; in direct chats the reply dispatcher drops replyToMessageId, so the send layer cannot use im.message.reply for P2P thread replies. (extensions/feishu/src/bot.ts:1279, ef58307f843e)
send-layer-already-supports-thread-reply: The Feishu send helper already calls client.im.message.reply and includes reply_in_thread: true when replyInThread is passed; the missing piece is the bot routing decision targeted by this PR. (extensions/feishu/src/send.ts:161, ef58307f843e)
current-test-gap: Current tests cover group thread forcing and many P2P paths, but the searched current tree has no P2P thread_id regression test like the one proposed by the PR. (extensions/feishu/src/bot.test.ts:2644, ef58307f843e)
pr-diff-security-review: Provided PR file context shows changes only to CHANGELOG.md, extensions/feishu/src/bot.ts, and extensions/feishu/src/bot.test.ts; it does not touch CI workflows, lockfiles, package metadata, install/build scripts, dependencies, downloaded artifacts, or secret-handling code. (e0a0ba6ce823)

Likely related people:

vincentkoc: Commented twice that ProjectClownfish pushed a narrow repair to keep the contributor path canonical, force-pushed the current PR head, and added/removed Clownfish review labels in the timeline. (role: recent branch repair and likely follow-up owner; confidence: high; commits: 4729e2e4121b, e0a0ba6ce823; files: CHANGELOG.md, extensions/feishu/src/bot.ts, extensions/feishu/src/bot.test.ts)
guoqunabc: Authored merged PR fix(feishu): comprehensive reply mechanism — outbound replyToId forwarding + topic-aware reply targeting #33789, which introduced the topic-aware Feishu reply-targeting behavior and the group-focused routing that the current P2P thread fix needs to extend without regressing group replies. (role: related prior behavior author; confidence: medium; commits: f61596311594; files: extensions/feishu/src/bot.ts, extensions/feishu/src/bot.test.ts)
Takhoffman: Merged the related Feishu reply-targeting PR fix(feishu): comprehensive reply mechanism — outbound replyToId forwarding + topic-aware reply targeting #33789 according to the prior review context; useful for historical routing context, though not necessarily the current follow-up owner. (role: prior PR merger; confidence: low; commits: 63ce7c74bdc0; files: extensions/feishu/src/bot.ts)

Remaining risk / open question:

No live Feishu P2P-thread smoke proof is recorded; the PR checklist still leaves manual verification open.
Functional regression risk is concentrated around preserving the group reply-target behavior from fix(feishu): comprehensive reply mechanism — outbound replyToId forwarding + topic-aware reply targeting #33789 while adding the P2P thread_id path.

Codex review notes: model gpt-5.5, reasoning high; reviewed against ef58307f843e.

vincentkoc · 2026-04-28T07:55:22Z

ProjectClownfish pushed a narrow repair to this branch so the original contributor path can stay canonical.

Source PR: #38808
Validation: pnpm check:changed
Contributor credit is preserved in the branch history and PR context.

When a user sends a message inside a Feishu thread (话题) in a P2P chat, the bot's reply was incorrectly sent outside the thread as a standalone message. This was caused by PR openclaw#33789 which fixed group reply targeting (openclaw#32980) but inadvertently broke P2P thread replies by gating all thread-aware logic behind isGroup checks. The fix distinguishes between two P2P scenarios based on the Feishu API contract (thread_id present = thread message, root_id only = quote reply): - P2P thread messages (thread_id present): reply to root_id with reply_in_thread=true so the response stays in the thread. - P2P plain replies (root_id only, no thread_id): reply to the triggering message as before. Group chat behavior is completely unchanged — the new condition (directThreadMessage) requires isDirect=true, which is always false for groups. Changes: - replyInThread: true for P2P thread messages (enables reply_in_thread API flag) - replyTargetMessageId: use root_id for P2P thread messages - threadReply: true for P2P thread messages (drives effectiveReplyInThread in dispatcher) - skipReplyToInMessages: false for P2P thread messages (so send layer uses im.message.reply instead of im.message.create) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

vincentkoc · 2026-04-28T18:28:33Z

ProjectClownfish pushed a narrow repair to this branch so the original contributor path can stay canonical.

Source PR: #38808
Validation: pnpm check:changed
Contributor credit is preserved in the branch history and PR context.

openclaw-clownfish · 2026-04-29T02:21:48Z

ProjectClownfish could not safely update this branch, so it opened a narrow replacement PR instead.

Replacement PR: #73972
Source PR: #38808
Contributor credit is preserved in the replacement PR body and changelog plan.

openclaw-barnacle Bot added channel: feishu Channel integration: feishu size: S labels Mar 7, 2026

xinhuagu mentioned this pull request Mar 8, 2026

fix(feishu): reply inside P2P threads instead of main chat #38814

Closed

openclaw-barnacle Bot added the stale Marked as stale due to inactivity label Apr 22, 2026

clawsweeper Bot mentioned this pull request Apr 26, 2026

[Bug]: bot replies outside thread in P2P direct message thread #38806

Open

openclaw-barnacle Bot removed the stale Marked as stale due to inactivity label Apr 26, 2026

clawsweeper Bot mentioned this pull request Apr 26, 2026

fix(feishu): reply inside P2P thread instead of as standalone message #38996

Closed

6 tasks

BingqingLyu mentioned this pull request Apr 27, 2026

fix(feishu): reply inside thread for P2P direct messages BingqingLyu/openclaw#64

Open

4 tasks

vincentkoc force-pushed the fix/feishu-p2p-thread-reply branch from 1bd3555 to 84d0830 Compare April 28, 2026 07:55

vincentkoc added clownfish:human-review labels Apr 28, 2026

LiaoyuanNing and others added 3 commits April 28, 2026 18:23

fix(feishu): reply inside P2P direct-message threads

4729e2e

fix(feishu): reply inside P2P direct-message threads

e0a0ba6

vincentkoc force-pushed the fix/feishu-p2p-thread-reply branch from 84d0830 to e0a0ba6 Compare April 28, 2026 18:28

vincentkoc added clawsweeper Tracked by ClawSweeper automation and removed clownfish:human-review labels Apr 28, 2026

openclaw-clownfish Bot mentioned this pull request Apr 29, 2026

fix(feishu): reply inside P2P direct-message threads #73972

Open

openclaw-clownfish Bot closed this Apr 29, 2026

openclaw-clownfish Bot mentioned this pull request Apr 29, 2026

fix(feishu): reply inside P2P direct-message threads #74001

Closed

clawsweeper Bot mentioned this pull request Apr 30, 2026

[Bug] Feishu P2P Topic - Bot cannot reply into thread, missing topic_id in webhook #74620

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

fix(feishu): reply inside thread for P2P direct messages#38808

fix(feishu): reply inside thread for P2P direct messages#38808
LiaoyuanNing wants to merge 3 commits intoopenclaw:mainfrom
LiaoyuanNing:fix/feishu-p2p-thread-reply

LiaoyuanNing commented Mar 7, 2026

Uh oh!

greptile-apps Bot commented Mar 7, 2026

Uh oh!

LiaoyuanNing commented Mar 9, 2026

Uh oh!

openclaw-barnacle Bot commented Apr 22, 2026

Uh oh!

clawsweeper Bot commented Apr 26, 2026 •

edited

Loading

Uh oh!

vincentkoc commented Apr 28, 2026

Uh oh!

vincentkoc commented Apr 28, 2026

Uh oh!

openclaw-clownfish Bot commented Apr 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

LiaoyuanNing commented Mar 7, 2026

Summary

Changes

Test plan

Uh oh!

greptile-apps Bot commented Mar 7, 2026

Greptile Summary

Confidence Score: 5/5

Uh oh!

LiaoyuanNing commented Mar 9, 2026

Uh oh!

openclaw-barnacle Bot commented Apr 22, 2026

Uh oh!

clawsweeper Bot commented Apr 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

vincentkoc commented Apr 28, 2026

Uh oh!

vincentkoc commented Apr 28, 2026

Uh oh!

openclaw-clownfish Bot commented Apr 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

clawsweeper Bot commented Apr 26, 2026 •

edited

Loading