fix: avoid replaying Responses item ids when store is disabled by zhanghang02 · Pull Request #84904 · openclaw/openclaw

zhanghang02 · 2026-05-21T09:40:17Z

Summary

Avoid replaying persisted OpenAI Responses item ids when the outgoing request disables store.
Keep stateless reasoning replay usable by preserving encrypted reasoning content and normalizing missing summary arrays.
Preserve replay item ids only when direct OpenAI store-enabled wrappers opt in.
Cover OpenAI Responses, ChatGPT/Codex Responses, GitHub Copilot Responses, and shared/simple Responses conversion paths.
Carry current-main channel ingress queue CI fallout fixes needed after rebase.

Why

When OpenClaw sends Responses requests with store: false, provider-owned item ids from a previous turn are not persisted. Replaying those ids in the next turn can make the provider reject the request with an item-not-found error because stored items are unavailable. Store-disabled paths should replay the useful content and encrypted reasoning state without replaying stale rs_*, msg_*, or fc_* item ids.

Real behavior proof

Behavior addressed: Follow-up Responses turns no longer replay server-owned reasoning/message/function-call item ids when the outgoing payload has store: false; store-enabled direct OpenAI wrappers can still opt into item-id replay.

Real environment tested: Local OpenClaw branch on Node 26 using live OpenAI Responses API (gpt-5.5) with OPENAI_API_KEY; local unit/type/lint checks on the rebased branch.

Exact steps or command run after this patch: Built a live stored setup response with reasoning + encrypted content, then sent a follow-up through streamOpenAIResponses itself. The captured outgoing payload had store:false, omitted prior reasoning/message/function-call item ids, preserved encrypted reasoning, normalized summary: [], and the live OpenAI request completed.

Evidence after fix: Redacted live proof from the direct builder request:

{"directPayload":{"store":false,"reasoningIdOmitted":true,"encryptedContentPreserved":true,"reasoningSummaryArray":true,"messageIdOmitted":true,"functionCallIdOmitted":true,"assistantPhase":"final_answer"},"liveResult":{"stopReason":"stop","contentTypes":["text"]}}

Additional live shared-converter proof completed with the same id omission and encrypted reasoning preservation, returning status:"completed" from OpenAI Responses.

Observed result after fix: The live follow-up request was accepted and completed. No raw provider ids or credentials were printed.

What was not tested: A live GitHub Copilot Responses request was not run from this machine; Copilot-specific behavior is covered by wrapper/unit regression tests for idless reasoning replay and oversized-id filtering.

Testing

node scripts/run-vitest.mjs extensions/github-copilot/connection-bound-ids.test.ts extensions/github-copilot/stream.test.ts src/llm/providers/openai-chatgpt-responses.test.ts src/llm/providers/openai-compatible-auth.test.ts src/llm/providers/openai-responses-shared.test.ts src/agents/openai-transport-stream.test.ts src/agents/embedded-agent-runner-extraparams.test.ts src/agents/openai-responses.reasoning-replay.test.ts
pnpm check:test-types
pnpm lint --threads=8
pnpm run lint:plugins:no-extension-test-core-imports
pnpm run lint:extensions:bundled
git diff --check origin/main...HEAD
git diff --check
/Users/steipete/Projects/agent-scripts/skills/autoreview/scripts/autoreview --mode branch --base origin/main -> clean, no accepted/actionable findings

clawsweeper · 2026-05-21T09:42:07Z

Codex review: needs maintainer review before merge. Reviewed May 31, 2026, 5:03 PM ET / 21:03 UTC.

Summary
The branch threads a replayResponsesItemIds policy through OpenAI Responses converters and wrappers so store-disabled payloads omit prior reasoning/message/function-call item IDs, with focused regressions and channel ingress test seam fixes.

PR surface: Source +75, Tests +619. Total +694 across 31 files.

Reproducibility: yes. Source inspection and reported live evidence show current main can build store:false Responses requests while replaying prior provider item IDs; I did not run a fresh live repro during this read-only review.

Review metrics: none identified.

Merge readiness
Overall: 🦞 diamond lobster
Proof: 🦞 diamond lobster
Patch quality: 🦞 diamond lobster
Result: ready for maintainer review.

Overall follows the weaker of proof and patch quality, so missing proof can cap an otherwise strong patch.

Risk before merge

[P1] Merging intentionally changes the replay contract for store-disabled Responses paths: existing setups that accidentally depended on provider-owned item IDs will now use idless stateless replay instead.
[P1] This prevents newly poisoned store-disabled turns, but already-poisoned long-lived sessions may still need reset or recovery work tracked by fix(agents): recover invalid OpenAI Responses reasoning replay #84267, Fix replay-invalid session recovery #85362, and openai-responses provider: 404 on previous_response_id when store=false (default) #88499.
[P1] Live proof covers OpenAI Responses; GitHub Copilot Responses behavior is covered by wrapper/unit regressions rather than a live Copilot run.
[P1] The latest head had CI restarted in the supplied context, so landing should wait for the current-head checks to finish cleanly.

Maintainer options:

Land after accepting the replay contract (recommended)
Maintainers can accept the intentional store-disabled idless replay behavior after latest-head CI is green, with the linked recovery work left as follow-up.
Require recovery before landing
Maintainers can pause this PR until the already-poisoned-session recovery path is merged if they want prevention and recovery to ship together.
Request live Copilot proof first
Maintainers can require a live GitHub Copilot Responses probe if unit-level Copilot coverage is not enough for the provider-specific part of the change.

Next step before merge

No automated repair is indicated; the next action is maintainer acceptance of the compatibility/session-state tradeoff plus latest-head CI verification.

Security
Cleared: No concrete security or supply-chain regression was found; the diff changes provider payload conversion and test seams, not secrets, workflows, dependencies, or install scripts.

Review details

Best possible solution:

Land the payload-layer prevention after maintainer acceptance of the replay-contract change and latest-head checks, while keeping already-poisoned session recovery in the linked follow-ups.

Do we have a high-confidence way to reproduce the issue?

Yes. Source inspection and reported live evidence show current main can build store:false Responses requests while replaying prior provider item IDs; I did not run a fresh live repro during this read-only review.

Is this the best way to solve the issue?

Yes. For the payload layer, deciding replay-ID policy before message conversion is the narrowest durable fix; retry/reset recovery for already-poisoned sessions is a separate layer tracked by related items.

AGENTS.md: found and applied where relevant.

Codex review notes: model gpt-5.5, reasoning high; reviewed against e5acae44534d.

Label changes

Label changes:

add rating: 🦞 diamond lobster: Overall readiness is 🦞 diamond lobster; proof is 🦞 diamond lobster and patch quality is 🦞 diamond lobster.
remove rating: 🐚 platinum hermit: Current PR rating is rating: 🦞 diamond lobster, so this older rating label is no longer current.

Label justifications:

P1: The PR targets an active OpenAI Responses replay failure that can break follow-up agent/channel turns for real users until session reset.
merge-risk: 🚨 compatibility: The diff intentionally changes the provider replay contract by omitting stored item IDs on store-disabled Responses requests.
merge-risk: 🚨 session-state: The affected state is conversation replay/session continuation state, and linked reports show stale replay data can poison long-lived sessions.
rating: 🦞 diamond lobster: Overall readiness is 🦞 diamond lobster; proof is 🦞 diamond lobster and patch quality is 🦞 diamond lobster.
status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (live_output): The PR body and maintainer comments include redacted live OpenAI Responses output showing after-fix store:false requests omit replay IDs, preserve encrypted reasoning, and complete successfully.
proof: sufficient: Contributor real behavior proof is sufficient. The PR body and maintainer comments include redacted live OpenAI Responses output showing after-fix store:false requests omit replay IDs, preserve encrypted reasoning, and complete successfully.

Evidence reviewed

PR surface:

Source +75, Tests +619. Total +694 across 31 files.

View PR surface stats

Area	Files	Added	Removed	Net
Source	14	125	50	+75
Tests	17	718	99	+619
Docs	0	0	0	0
Config	0	0	0	0
Generated	0	0	0	0
Other	0	0	0	0
Total	31	843	149	+694

What I checked:

Current main still sends store-disabled OpenAI Responses payloads through the default replay-id path: On current main, streamOpenAIResponses builds store: false payloads while calling convertResponsesMessages without a replay-id override, so current main does not yet implement the broader store-disabled item-id gate. (src/llm/providers/openai-responses.ts:188, e5acae44534d)
PR gates replay IDs before building the Responses input: The branch resolves the store policy before conversion and passes replayResponsesItemIds into convertResponsesMessages; store-disabled native paths therefore convert replay items without provider-owned IDs. (src/agents/openai-transport-stream.ts:2094, 20b9a19b5634)
PR deletes prior item IDs while keeping stateless reasoning material: The changed converter normalizes missing reasoning summaries to [], deletes reasoning IDs when replay IDs are disabled, omits assistant message IDs, and omits function-call item IDs while retaining call_id. (src/llm/providers/openai-responses-shared.ts:46, 20b9a19b5634)
Focused regression coverage locks the store-disabled behavior: The branch adds tests that build a store-disabled Responses payload and assert the reasoning, assistant message, and function-call replay IDs are omitted while encrypted reasoning and summary arrays remain. (src/agents/openai-transport-stream.test.ts:2781, 20b9a19b5634)
OpenAI dependency contract supports stateless encrypted reasoning replay: The OpenAI Node 6.39.1 Responses types document reasoning.encrypted_content as enabling reasoning items in stateless multi-turn conversations when store is false, and expose store, previous_response_id, reasoning summaries, and encrypted content on the relevant request/item shapes.
Codex sibling source was inspected for Responses item shape compatibility: The sibling Codex source models reasoning response items with required summary, optional encrypted content, and function-call IDs as optional, which is compatible with idless stateless replay for this OpenClaw payload layer. (../codex/sdk/python/src/openai_codex/generated/v2_all.py:2822, e93dc98a48d5)

Likely related people:

Ben Newell: Blame on current main shows the existing store-disabled Responses builder and converter scaffolding are from commit a88e4fb. (role: introduced current main Responses transport/converter surface; confidence: high; commits: a88e4fb7e05e; files: src/llm/providers/openai-responses.ts, src/agents/openai-transport-stream.ts, src/llm/providers/openai-responses-shared.ts)
Simon Peck: The merged adjacent replay-id policy in commit 6653193 owns the current main paired message-id logic this PR builds on. (role: recent adjacent replay-policy contributor; confidence: high; commits: 6653193fdb90; files: src/agents/openai-transport-stream.ts, src/llm/providers/openai-responses-shared.ts)
steipete: The provided PR timeline shows maintainer fixups, proof refreshes, and re-review requests on the current PR head, making this a practical follow-up routing candidate even though the source history predates the branch fixups. (role: recent branch fixer and reviewer; confidence: medium; commits: 7de9fee5fbfd, e6bf803aa757, b0cfe6861bce; files: src/agents/openai-transport-stream.ts, src/llm/providers/openai-responses-shared.ts, src/llm/providers/openai-chatgpt-responses.ts)

What the crustacean ranks mean

🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

How this review workflow works

ClawSweeper keeps one durable marker-backed review comment per issue or PR.
Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
Maintainers can also comment @clawsweeper review to request a fresh review only.
Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

clawsweeper · 2026-05-21T09:49:20Z

ClawSweeper PR egg

✨ Hatched: 🥚 common Brave Crabkin

Hatch command

Comment @clawsweeper hatch when this PR is hatchable.

Hatchability rules:

Merged PRs are hatchable.
Open PRs are hatchable when they are status: 👀 ready for maintainer look, status: 🚀 automerge armed, or labeled clawsweeper:automerge.
Closed unmerged PRs are hatchable only when one of those hatchable labels is still present in the durable record.

Rarity: 🥚 common.
Trait: keeps receipts.
Image traits: location status garden; accessory review stamp; palette violet, aqua, and starlight; mood sleepy but ready; pose balancing on a branch marker; shell brushed metal shell; lighting soft underwater shimmer; background small review tokens.
Share on X: post this hatch
Copy: My PR egg hatched a 🥚 common Brave Crabkin in ClawSweeper.

What is this egg doing here?

Eggs appear after the PR passes real-behavior proof. It is here for vibes, not verdicts: it does not change labels, ratings, merge decisions, or automation.
The shell reacts to review momentum: open follow-up work warms it up, re-review makes it wobble, and a clean final review lets it hatch.
Hatchability usually comes from sufficient real-behavior proof, no blocking P0/P1/P2 findings, no security attention needed, and clean correctness. A merged PR is already final, so merge makes the egg hatchable independently.
The hatch is seeded from this repository and PR number, so the same PR keeps the same creature; the reviewed head SHA can only change safe visual details.
Rarity is just collectible sparkle: 🥚 common, 🌱 uncommon, 💎 rare, ✨ glimmer, and 🌈 legendary.

zhanghang02 · 2026-05-22T01:41:35Z

@clawsweeper re-review

clawsweeper · 2026-05-22T01:43:20Z

🦞🧹
ClawSweeper re-review requested.

I asked ClawSweeper to review this item again.
Action: item re-review queued (workflow sweep.yml, event repository_dispatch).
Result: the existing ClawSweeper review comment will be edited in place when the review finishes.

latensified · 2026-05-22T19:17:05Z

Cross-linking the Responses item-id coordination note from #85277:
#85277 (comment)

For this PR, the proposed final layering is that the store policy gate wins first: when the final Responses payload is store:false, omit all provider-owned item ids (reasoning.id, assistant message.id, and function_call.id). When a wrapper explicitly sends store:true, id replay can remain enabled and the per-item safety rules from #85277 decide whether an assistant message.id is still replayable.

Suggested merge order from the coordination pass: #84904 -> #85277 -> #84267 -> #85362. That keeps this PR as the broad persistence-contract gate, then lets the signed-message pairing rule and retry/reset behavior layer on top.

tianxiaochannel-oss88 · 2026-05-31T03:07:27Z

This PR still looks like the closest source-level fix for a fresh 2026.5.28 macOS/Slack incident I hit today.

Live evidence from the local gateway:

OpenClaw CLI/Gateway: 2026.5.28 (e932160)
Channel: Slack channel session
Provider/API path: openai / openai-responses
Model: gpt-5.5
Slack health stayed OK: running, connected, health:healthy, works
Gateway health stayed OK: Connectivity probe: ok
The assistant turn failed before reply with:

HTTP 404: Item with id 'rs_[redacted]' not found. Items are not persisted when `store` is set to false. Try again with `store` set to true, or remove this item from your input.

I recovered the local install with a two-layer workaround:

Reset the affected Slack channel session to clear the already-poisoned replay state.
Added a local transport retry guard that treats thinking_signature_invalid and Item with id ... rs_... not found / store=false as replay-invalid, strips stale Responses replay state (reasoning items, encrypted_content, and provider-owned replay ids), and retries once.

That local installed-dist workaround is not a proposed upstream shape by itself; it is just operational proof that store-disabled replay ids are the failing boundary. The source-level fix in this PR is the better durable direction because it prevents store:false requests from sending provider-owned item ids in the first place, while preserving store-enabled replay where appropriate.

I would still expect a separate session-recovery layer to be useful for already-poisoned long-lived channel sessions, but this PR appears to cover the transport/payload layer that prevents new poisoned turns.

steipete · 2026-05-31T20:56:34Z

Force-pushed one more lint-only fixup at 20b9a19.

Delta since the previous proof comment:

avoided the CI-only await lint in the WhatsApp runtime mocks.
fixed bundled Feishu runtime type exports that lint:extensions:bundled exposed.

Fresh proof after this fixup:

pnpm run lint:extensions:bundled
pnpm lint --threads=8
pnpm check:test-types
git diff --check and git diff --check origin/main...HEAD
autoreview branch pass against current origin/main: clean, no accepted/actionable findings.

CI restarted on 20b9a19; watching it now.

steipete · 2026-05-31T21:12:21Z

Rebased on latest origin/main again and force-pushed 5f3e7af.

Delta since the last comment:

fixed the merged-base task boundary failure by routing async task runner lookups through task-status-access instead of importing task-registry directly.
fixed the remaining Feishu bundled lint await warning.

Fresh proof after this rebase:

node scripts/run-vitest.mjs extensions/github-copilot/connection-bound-ids.test.ts extensions/github-copilot/stream.test.ts src/llm/providers/openai-chatgpt-responses.test.ts src/llm/providers/openai-compatible-auth.test.ts src/llm/providers/openai-responses-shared.test.ts src/agents/openai-transport-stream.test.ts src/agents/embedded-agent-runner-extraparams.test.ts src/agents/openai-responses.reasoning-replay.test.ts src/tasks/task-boundaries.test.ts
pnpm run lint:extensions:bundled
pnpm lint --threads=8
pnpm check:test-types
git diff --check and git diff --check origin/main...HEAD
autoreview branch pass against current origin/main: clean, no accepted/actionable findings.

CI restarted on 5f3e7af; watching it now.

steipete · 2026-05-31T21:17:15Z

Land-ready on 5f3e7af.

Local proof:

node scripts/run-vitest.mjs extensions/github-copilot/connection-bound-ids.test.ts extensions/github-copilot/stream.test.ts src/llm/providers/openai-chatgpt-responses.test.ts src/llm/providers/openai-compatible-auth.test.ts src/llm/providers/openai-responses-shared.test.ts src/agents/openai-transport-stream.test.ts src/agents/embedded-agent-runner-extraparams.test.ts src/agents/openai-responses.reasoning-replay.test.ts src/tasks/task-boundaries.test.ts
pnpm run lint:extensions:bundled
pnpm lint --threads=8
pnpm check:test-types
pnpm run lint:plugins:no-extension-test-core-imports
git diff --check and git diff --check origin/main...HEAD
live OpenAI Responses gpt-5.5 through streamOpenAIResponses: store:false, reasoning/message/function-call ids omitted, encrypted reasoning preserved, summary array present, request completed with stopReason=stop.
autoreview branch pass against current origin/main: clean, no accepted/actionable findings.

CI proof:

GitHub PR checks are clean for 5f3e7af. The remaining canceled Real behavior proof entry is from a superseded force-push, not the current head.

Known gap:

No live GitHub Copilot Responses request from this machine; Copilot-specific behavior is covered by unit regression tests for idless reasoning replay and oversized-id filtering.

openclaw-barnacle Bot added agents Agent runtime and tooling size: S triage: needs-real-behavior-proof Candidate: external PR needs after-fix proof from a real setup. labels May 21, 2026

zhanghang02 force-pushed the fix/responses-store-false-replay-ids branch from 487490e to c4c36ac Compare May 22, 2026 01:41

openclaw-barnacle Bot added size: M and removed size: S labels May 22, 2026

openclaw-barnacle Bot added proof: supplied External PR includes structured after-fix real behavior proof. and removed triage: needs-real-behavior-proof Candidate: external PR needs after-fix proof from a real setup. labels May 22, 2026

clawsweeper Bot mentioned this pull request May 22, 2026

fix(openai): avoid stale Responses message id replay #85277

Merged

25 tasks

This was referenced May 22, 2026

Fix replay-invalid session recovery #85362

Closed

fix(agents): recover invalid OpenAI Responses reasoning replay #84267

Closed

This was referenced May 31, 2026

openai-responses provider: 404 on previous_response_id when store=false (default) #88499

Closed

fix(agents): recover stale OpenAI Responses continuation state #86817

Closed

steipete self-assigned this May 31, 2026

openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 31, 2026

clawsweeper Bot added the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 31, 2026

openclaw-barnacle Bot added the channel: feishu Channel integration: feishu label May 31, 2026

zhanghang02 and others added 13 commits May 31, 2026 22:02

fix: avoid replaying Responses item ids when store is disabled

1598a62

fix(openai): keep replay reasoning summary

ad856b7

fix(openai): honor store replay policy in shared responses

c192c69

fix(openai): preserve encrypted reasoning without replay ids

df3e7f3

fix(openai): default shared responses replay ids off

8d9f2f6

fix(openai): disable chatgpt response id replay

4d561f8

fix(openai): keep copilot reasoning without replay ids

effafe9

fix(copilot): allow idless reasoning replay

533a9c5

test(openai): fix responses replay test types

9be5e46

test(copilot): keep idless reasoning replay expectation

f79930f

test(channels): fix ingress queue CI fallout

9fd8b34

test(extensions): fix bundled lint fallout

02d14d9

fix(feishu): satisfy bundled reply lint

851a5e7

clawsweeper Bot added rating: 🦞 diamond lobster Very strong PR readiness with only minor maintainer review expected. and removed rating: 🐚 platinum hermit Good normal PR readiness with ordinary maintainer review expected. labels May 31, 2026

fix(tasks): route async task lookup through status access

5f3e7af

steipete force-pushed the fix/responses-store-false-replay-ids branch from 20b9a19 to 5f3e7af Compare May 31, 2026 21:12

openclaw-barnacle Bot added size: XL and removed size: L labels May 31, 2026

steipete merged commit 03dec8b into openclaw:main May 31, 2026
166 of 168 checks passed

toruvieI mentioned this pull request Jun 4, 2026

openai-chatgpt-responses native replay sends encrypted reasoning and breaks next turn with invalid_encrypted_content #90093

Closed

This was referenced Jun 15, 2026

fix(check): clean ingress queue lint blockers #88780

Closed

Prune completed Codex Responses tool replay #79655

Open

fix(github-copilot): strip encrypted_content from reasoning replay items #95493

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

fix: avoid replaying Responses item ids when store is disabled#84904

fix: avoid replaying Responses item ids when store is disabled#84904
steipete merged 14 commits into
openclaw:mainfrom
zhanghang02:fix/responses-store-false-replay-ids

zhanghang02 commented May 21, 2026 •

edited by steipete

Loading

Uh oh!

clawsweeper Bot commented May 21, 2026 •

edited

Loading

Uh oh!

clawsweeper Bot commented May 21, 2026 •

edited

Loading

Uh oh!

zhanghang02 commented May 22, 2026

Uh oh!

clawsweeper Bot commented May 22, 2026

Uh oh!

latensified commented May 22, 2026

Uh oh!

tianxiaochannel-oss88 commented May 31, 2026

Uh oh!

steipete commented May 31, 2026

Uh oh!

steipete commented May 31, 2026

Uh oh!

steipete commented May 31, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Uh oh!

Uh oh!

Conversation

zhanghang02 commented May 21, 2026 • edited by steipete Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Why

Real behavior proof

Testing

Uh oh!

clawsweeper Bot commented May 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

clawsweeper Bot commented May 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Hatch command

Uh oh!

zhanghang02 commented May 22, 2026

Uh oh!

clawsweeper Bot commented May 22, 2026

Uh oh!

latensified commented May 22, 2026

Uh oh!

tianxiaochannel-oss88 commented May 31, 2026

Uh oh!

steipete commented May 31, 2026

Uh oh!

steipete commented May 31, 2026

Uh oh!

steipete commented May 31, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

zhanghang02 commented May 21, 2026 •

edited by steipete

Loading

clawsweeper Bot commented May 21, 2026 •

edited

Loading

clawsweeper Bot commented May 21, 2026 •

edited

Loading