repo stringclasses 1
value | github_id int64 3.77B 4.38B | github_node_id stringlengths 18 24 | number int64 75 77.8k | html_url stringlengths 46 49 | api_url stringlengths 56 59 | title stringlengths 1 746 | body stringlengths 1 55.7k ⌀ | state stringclasses 2
values | state_reason stringclasses 4
values | locked bool 2
classes | comments_count int64 0 128 | labels listlengths 0 5 | assignees listlengths 0 2 | created_at stringdate 2026-01-01 16:55:34 2026-05-05 11:59:03 | updated_at stringdate 2026-03-24 00:13:29 2026-05-05 11:59:15 | closed_at stringlengths 20 20 ⌀ | author_association stringclasses 3
values | milestone_title stringclasses 0
values | snapshot_id stringclasses 76
values | extracted_at stringdate 2026-03-24 23:36:49 2026-05-05 12:00:33 | author_login stringlengths 3 38 | author_id int64 346 282M | author_node_id stringlengths 12 20 | author_type stringclasses 2
values | author_site_admin bool 2
classes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
openclaw/openclaw | 4,272,396,363 | I_kwDOQb6kR87-p5hL | 67,435 | https://github.com/openclaw/openclaw/issues/67435 | https://api.github.com/repos/openclaw/openclaw/issues/67435 | [Bug]: Ollama provider prepends `ollama/` prefix to model ID, causing 404 from Ollama API | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
## Summary
When using the `ollama` provider, OpenClaw constructs the model string as `ollama/<modelId>` before sending it to the Ollama API. Ollama does not accept this prefix — it expects the bare model name (e.g., `qwen3:... | closed | completed | true | 4 | [
"bug",
"regression"
] | [] | 2026-04-15T23:56:14Z | 2026-04-19T04:22:51Z | 2026-04-16T05:45:40Z | NONE | null | 20260419T052035Z | 2026-04-19T05:20:35Z | StevoKeano | 935,091 | MDQ6VXNlcjkzNTA5MQ== | User | false |
openclaw/openclaw | 4,272,397,175 | I_kwDOQb6kR87-p5t3 | 67,436 | https://github.com/openclaw/openclaw/issues/67436 | https://api.github.com/repos/openclaw/openclaw/issues/67436 | [Bug]: Gateway self-induced hot-reload loop corrupts manifest.db on Linux VPS | ### Bug type
Crash (process/app exits or hangs)
### Beta release blocker
No
### Summary
On Linux VPS with systemd, gateway auto-provisions plugins on boot (brave, anthropic, openai, memory-core) by writing to openclaw.json via atomic rename, which triggers its own hot-reload watcher → SIGUSR1 every 3-6 min → kills... | closed | completed | true | 1 | [
"bug"
] | [] | 2026-04-15T23:56:30Z | 2026-04-19T04:22:58Z | 2026-04-16T13:18:27Z | NONE | null | 20260419T052035Z | 2026-04-19T05:20:35Z | leythlahcene-max | 268,696,524 | U_kgDOEAP7zA | User | false |
openclaw/openclaw | 4,272,429,674 | I_kwDOQb6kR87-qBpq | 67,438 | https://github.com/openclaw/openclaw/issues/67438 | https://api.github.com/repos/openclaw/openclaw/issues/67438 | [Bug]: FeiShu QR Code is not woking when using the openclaw onboard | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Using OpenClaw 2026.4.14 (323493f)
FeiShu QR Code is not woking when using the openclaw onboard
### Steps to reproduce
When I try to onboard the openclaw with 'openclaw onboard' in cmd window, then try to se... | open | null | false | 3 | [
"bug",
"bug:behavior"
] | [] | 2026-04-16T00:09:40Z | 2026-05-02T21:49:54Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | 381426068 | 10,822,009 | MDQ6VXNlcjEwODIyMDA5 | User | false |
openclaw/openclaw | 4,272,438,965 | I_kwDOQb6kR87-qD61 | 67,439 | https://github.com/openclaw/openclaw/issues/67439 | https://api.github.com/repos/openclaw/openclaw/issues/67439 | Cron CLI commands extremely slow (~20s) and exit abnormally after successful mutations | ## Summary
`openclaw cron add/edit/rm/list` commands consistently take 15-20+ seconds to complete and often exit with SIGKILL/SIGTERM signals even when the underlying mutation succeeds. The cron engine and jobs.json storage layer are unaffected — only the CLI/RPC path is degraded.
## Environment
- OpenClaw 2026.4.11... | closed | completed | true | 4 | [] | [] | 2026-04-16T00:13:21Z | 2026-04-28T04:53:10Z | 2026-04-25T18:08:11Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | meli-clawbot | 262,196,076 | U_kgDOD6DLbA | User | false |
openclaw/openclaw | 4,272,443,766 | I_kwDOQb6kR87-qFF2 | 67,440 | https://github.com/openclaw/openclaw/issues/67440 | https://api.github.com/repos/openclaw/openclaw/issues/67440 | [Feature][Security]: Add optional TOTP (authenticator app code) to exec approvals | ### Summary
Add an optional TOTP challenge (6-digit authenticator code) to exec approvals. Right now, /approve is just a string any authorized sender can type, there's nothing verifying it's actually you. The default config skips the whole approval step entirely
I already use TOTP via `pam_oath` for sudo on my headle... | open | null | false | 4 | [
"enhancement",
"security"
] | [] | 2026-04-16T00:15:01Z | 2026-04-17T23:18:34Z | null | NONE | null | 20260417T232041Z | 2026-04-17T23:20:41Z | tahoe-ai | 166,661,269 | U_kgDOCe8MlQ | User | false |
openclaw/openclaw | 4,272,453,971 | I_kwDOQb6kR87-qHlT | 67,441 | https://github.com/openclaw/openclaw/issues/67441 | https://api.github.com/repos/openclaw/openclaw/issues/67441 | Cron job reports lastRunStatus=error when delivery succeeded (false positive from QA transport) | ## Bug Description
When a cron job runs an isolated agent session that posts to Discord via the `message` tool and returns `NO_REPLY`, the cron runner marks the job as `lastRunStatus: "error"` with `lastError: "⚠️ ✉️ Message failed"` — even though `lastDelivered: true` and `lastDeliveryStatus: "delivered"` are both se... | closed | completed | true | 2 | [] | [] | 2026-04-16T00:18:58Z | 2026-04-28T04:53:09Z | 2026-04-25T18:08:18Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | ludovicfourrage | 28,280,112 | MDQ6VXNlcjI4MjgwMTEy | User | false |
openclaw/openclaw | 4,272,455,168 | I_kwDOQb6kR87-qH4A | 67,442 | https://github.com/openclaw/openclaw/issues/67442 | https://api.github.com/repos/openclaw/openclaw/issues/67442 | [Bug]: Dreaming can promote transport/session metadata into MEMORY.md as durable memory | ### Bug type
Behavior bug (incorrect output quality)
### Summary
Dreaming can stage raw transport/session metadata as durable short-term candidates and later promote that garbage into `MEMORY.md`.
This is more severe than the already-reported "daily file pollution" / "raw session-corpus reflections" issues, because... | closed | completed | true | 2 | [] | [] | 2026-04-16T00:19:26Z | 2026-04-28T04:53:08Z | 2026-04-25T18:08:25Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Brisbanehuang | 202,969,697 | U_kgDODBkSYQ | User | false |
openclaw/openclaw | 4,272,484,237 | I_kwDOQb6kR87-qO-N | 67,445 | https://github.com/openclaw/openclaw/issues/67445 | https://api.github.com/repos/openclaw/openclaw/issues/67445 | [Feature]: Explicit agent mentions should override channel-default auto-responder | ## Problem
In channels where one agent is configured to auto-respond to every message, that default behavior becomes noisy when a user occasionally wants to address a different agent directly.
Today, the channel-default agent still responds, even when the message explicitly tags another agent.
## Desired behavior
W... | closed | not_planned | true | 1 | [] | [] | 2026-04-16T00:30:29Z | 2026-04-30T04:39:54Z | 2026-04-26T05:12:56Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | unboxed-ai | 253,808,488 | U_kgDODyDPaA | User | false |
openclaw/openclaw | 4,272,553,870 | I_kwDOQb6kR87-qf-O | 67,453 | https://github.com/openclaw/openclaw/issues/67453 | https://api.github.com/repos/openclaw/openclaw/issues/67453 | Discord voice joins and generates replies, but playback fails: microsoft TTS provider not registered | ## Summary
Discord voice can successfully join, capture speech, transcribe audio, and generate reply text, but fails at the playback stage because no usable TTS provider is registered in the runtime.
This appears to be distinct from the earlier `/vc join` handshake-stage failure tracked in #65039.
## Environment
- ... | closed | completed | true | 1 | [] | [] | 2026-04-16T00:54:18Z | 2026-04-27T04:34:14Z | 2026-04-25T04:06:20Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | dyerrington | 2,257,834 | MDQ6VXNlcjIyNTc4MzQ= | User | false |
openclaw/openclaw | 4,272,557,901 | I_kwDOQb6kR87-qg9N | 67,454 | https://github.com/openclaw/openclaw/issues/67454 | https://api.github.com/repos/openclaw/openclaw/issues/67454 | [Bug] OpenRouter returns empty assistant payloads in Telegram and Control UI on macOS 26.2 | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
With OpenRouter configured, assistant responses are empty in both Telegram and the Control UI, and gateway logs show `incomplete turn detected ... payloads=0`.
### Steps to reproduce
1. Install OpenClaw 2026.... | closed | completed | true | 4 | [
"bug",
"bug:behavior"
] | [] | 2026-04-16T00:55:59Z | 2026-04-28T04:53:07Z | 2026-04-25T18:08:47Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | ddlee12 | 85,885,770 | MDQ6VXNlcjg1ODg1Nzcw | User | false |
openclaw/openclaw | 4,272,598,519 | I_kwDOQb6kR87-qq33 | 67,458 | https://github.com/openclaw/openclaw/issues/67458 | https://api.github.com/repos/openclaw/openclaw/issues/67458 | [Feature] Gemini 3.1 Flash TTS support in messages.tts, with expressive tag passthrough | OpenClaw already has a nice speech-provider architecture behind `messages.tts` and supports providers like ElevenLabs, OpenAI, Microsoft, and MiniMax. It would be great to add Gemini 3.1 Flash TTS as a first-class speech provider.
## Why this would be valuable
Gemini 3.1 Flash TTS appears to support expressive inline ... | closed | completed | true | 0 | [] | [] | 2026-04-16T01:10:17Z | 2026-04-19T04:23:25Z | 2026-04-16T06:24:37Z | CONTRIBUTOR | null | 20260419T052035Z | 2026-04-19T05:20:35Z | barronlroth | 5,567,884 | MDQ6VXNlcjU1Njc4ODQ= | User | false |
openclaw/openclaw | 4,272,616,459 | I_kwDOQb6kR87-qvQL | 67,459 | https://github.com/openclaw/openclaw/issues/67459 | https://api.github.com/repos/openclaw/openclaw/issues/67459 | Skills snapshot cache not invalidated when new skills are added | ## Bug
When new skills are installed to any of the skill directories (`~/.openclaw/skills/`, `~/.agents/skills/`, workspace `.agents/skills/`, etc.), existing agent sessions do not pick them up. The agent's system prompt continues using a stale skill list from when the session was first created.
## Root Cause
Two ca... | closed | completed | true | 1 | [] | [] | 2026-04-16T01:17:05Z | 2026-04-28T04:53:07Z | 2026-04-25T18:08:55Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | LanicBlue | 47,910,958 | MDQ6VXNlcjQ3OTEwOTU4 | User | false |
openclaw/openclaw | 4,272,620,397 | I_kwDOQb6kR87-qwNt | 67,461 | https://github.com/openclaw/openclaw/issues/67461 | https://api.github.com/repos/openclaw/openclaw/issues/67461 | Gateway leaks undici sockets on every streamed Anthropic API call (buildManagedResponse missing finalize on GC) | ## Summary
The openclaw-gateway process accumulates leaked IPv6 ESTABLISHED sockets to api.anthropic.com (via Cloudflare). After 5 hours of normal cron-driven agent-turn executions, we observed 11,507 leaked ESTABLISHED IPv6 sockets, eventually causing `spawn EBADF` errors when the process hits its fd limit. All `exec... | open | null | false | 2 | [] | [] | 2026-04-16T01:18:33Z | 2026-05-03T08:25:23Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | jakedwyer | 1,486,207 | MDQ6VXNlcjE0ODYyMDc= | User | false |
openclaw/openclaw | 4,272,650,046 | I_kwDOQb6kR87-q3c- | 67,465 | https://github.com/openclaw/openclaw/issues/67465 | https://api.github.com/repos/openclaw/openclaw/issues/67465 | Control UI/WebChat: configured Talk mode has no visible UI surface; web voice still appears browser-only | ### Summary
Control UI/WebChat exposes browser speech controls, but does not expose a usable Talk-mode UI even when `talk.*` is configured and loaded by the gateway. This creates a confusing mismatch: gateway Talk config appears valid, but the web UI still only offers browser-dependent voice pieces.
### What we obser... | closed | completed | true | 1 | [] | [] | 2026-04-16T01:29:52Z | 2026-04-28T04:53:06Z | 2026-04-25T18:09:02Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | WolvenRA | 35,547,932 | MDQ6VXNlcjM1NTQ3OTMy | User | false |
openclaw/openclaw | 4,272,655,479 | I_kwDOQb6kR87-q4x3 | 67,469 | https://github.com/openclaw/openclaw/issues/67469 | https://api.github.com/repos/openclaw/openclaw/issues/67469 | [Bug]: google-gemini-cli replies stop being persisted to OpenClaw session transcripts / WebUI history | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Discord chat link from Krill:
https://discord.com/channels/1456350064065904867/1494135269748772864
### Summary
When using `google-gemini-cli/*` models in the OpenClaw UI, new assistant messages stop appearing in OpenClaw session h... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-16T01:32:00Z | 2026-04-18T04:14:05Z | 2026-04-16T04:00:34Z | NONE | null | 20260418T042035Z | 2026-04-18T04:20:35Z | Trung0246 | 11,626,920 | MDQ6VXNlcjExNjI2OTIw | User | false |
openclaw/openclaw | 4,272,716,433 | I_kwDOQb6kR87-rHqR | 67,473 | https://github.com/openclaw/openclaw/issues/67473 | https://api.github.com/repos/openclaw/openclaw/issues/67473 | [Bug]: Auto-compaction (threshold-based) never fires in embedded runner | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Pi's threshold-based auto-compaction never fires in the embedded runner. Sessions grow past the configured trigger without compacting. Only manual `/compact` works. Confirmed by session JSONL — all compaction e... | closed | completed | true | 1 | [
"bug",
"bug:behavior"
] | [] | 2026-04-16T01:52:41Z | 2026-04-19T04:23:29Z | 2026-04-16T04:32:48Z | NONE | null | 20260419T052035Z | 2026-04-19T05:20:35Z | natechicago | 2,934,380 | MDQ6VXNlcjI5MzQzODA= | User | false |
openclaw/openclaw | 4,272,826,299 | I_kwDOQb6kR87-rie7 | 67,478 | https://github.com/openclaw/openclaw/issues/67478 | https://api.github.com/repos/openclaw/openclaw/issues/67478 | [Bug]: 1-Click Droplet has no guard against running openclaw as root — causes state corruption, crash loop, and backup failures | ### Bug type
Crash (process/app exits or hangs)
### Beta release blocker
No
### Summary
The DigitalOcean 1-Click Droplet image provides no guard against invoking `openclaw` directly as root. Since root is the default SSH user, running any `openclaw` subcommand as root (e.g. `openclaw doctor --fix`) silently corrup... | open | null | false | 3 | [
"bug",
"bug:crash"
] | [] | 2026-04-16T02:21:19Z | 2026-05-02T23:42:51Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | natechicago | 2,934,380 | MDQ6VXNlcjI5MzQzODA= | User | false |
openclaw/openclaw | 4,272,833,618 | I_kwDOQb6kR87-rkRS | 67,482 | https://github.com/openclaw/openclaw/issues/67482 | https://api.github.com/repos/openclaw/openclaw/issues/67482 | [Bug]: TypeError: Cannot read properties of undefined (reading 'trim') during init wizard | ### Bug type
Crash (process/app exits or hangs)
### Beta release blocker
No
### Summary
A TypeError occurs during the initialization wizard when reaching the "Select channel (QuickStart)" step. The interface displays several undefined: undefined entries in the channel status list before crashing with Cannot read p... | closed | completed | true | 9 | [
"bug",
"bug:crash"
] | [] | 2026-04-16T02:23:12Z | 2026-04-19T04:23:13Z | 2026-04-16T06:23:57Z | NONE | null | 20260419T052035Z | 2026-04-19T05:20:35Z | Cikian | 89,137,866 | MDQ6VXNlcjg5MTM3ODY2 | User | false |
openclaw/openclaw | 4,272,843,073 | I_kwDOQb6kR87-rmlB | 67,483 | https://github.com/openclaw/openclaw/issues/67483 | https://api.github.com/repos/openclaw/openclaw/issues/67483 | TypeError in formatDocsLink when channel plugin meta.docsPath is undefined during onboarding | ## Bug Description
During onboarding (`openclaw onboard`), selecting a configured channel and choosing "Skip (leave as-is)" causes a crash:
```
TypeError: Cannot read properties of undefined (reading 'trim')
```
## Steps to Reproduce
1. Run `openclaw onboard` (quickstart flow)
2. Have a channel already configured (... | closed | completed | true | 1 | [] | [] | 2026-04-16T02:25:41Z | 2026-04-18T04:14:10Z | 2026-04-16T02:41:18Z | NONE | null | 20260418T042035Z | 2026-04-18T04:20:35Z | ruiwang20010702 | 224,701,776 | U_kgDODWStUA | User | false |
openclaw/openclaw | 4,272,874,100 | I_kwDOQb6kR87-ruJ0 | 67,485 | https://github.com/openclaw/openclaw/issues/67485 | https://api.github.com/repos/openclaw/openclaw/issues/67485 | openclaw update loses bin symlink in nvm environment | ## Problem
When using `openclaw update` in an nvm-managed Node environment, the `openclaw` command becomes unavailable after the update completes.
## Steps to reproduce
1. Install openclaw with nvm on a specific Node version (e.g., `nvm use v16 && npm install -g openclaw`)
2. Run `openclaw update` to update to the l... | closed | completed | true | 1 | [] | [] | 2026-04-16T02:33:56Z | 2026-04-28T04:53:05Z | 2026-04-25T18:09:32Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | taopengpeng | 24,906,366 | MDQ6VXNlcjI0OTA2MzY2 | User | false |
openclaw/openclaw | 4,272,911,942 | I_kwDOQb6kR87-r3ZG | 67,486 | https://github.com/openclaw/openclaw/issues/67486 | https://api.github.com/repos/openclaw/openclaw/issues/67486 | [BlueBubbles] Hardcoded 10s send timeout causes message loss on macOS 26 (Private API stalls) | ### Summary
The BlueBubbles channel plugin uses a hardcoded 10-second timeout (`DEFAULT_TIMEOUT_MS = 1e4`) for all HTTP requests, including message sends via `/api/v1/message/text`. On macOS 26 (Tahoe), BlueBubbles Private API sends intermittently stall for 60-124 seconds due to iMessage framework delays. The aggressi... | closed | completed | true | 1 | [
"channel: bluebubbles"
] | [
"omarshahine"
] | 2026-04-16T02:44:39Z | 2026-04-23T04:22:27Z | 2026-04-20T17:04:54Z | NONE | null | 20260423T060020Z | 2026-04-23T06:00:20Z | larrylhollan | 265,963,893 | U_kgDOD9pJdQ | User | false |
openclaw/openclaw | 4,272,918,791 | I_kwDOQb6kR87-r5EH | 67,487 | https://github.com/openclaw/openclaw/issues/67487 | https://api.github.com/repos/openclaw/openclaw/issues/67487 | [Bug]: Control UI 图片上传后不显示在对话框 | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
**Bug 标题：**
Control UI 图片上传后不显示在对话框
**环境：**
- OpenClaw 版本：2026.4.9
- 浏览器：Chrome（最新版）
- 访问方式：本地 webchat（`http://127.0.0.1:18789/`）
**复现步骤：**
1. 打开 Control UI 聊天界面
2. 尝试粘贴或上传一张图片
3. 看到图片上传进度完成
4. 图片未出现在对话框消息气泡中... | closed | completed | true | 1 | [
"bug",
"close:duplicate",
"close:already-fixed",
"dedupe:child",
"bug:behavior"
] | [
"vincentkoc"
] | 2026-04-16T02:46:36Z | 2026-04-30T04:47:20Z | 2026-04-25T06:08:12Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | laypan1 | 259,503,309 | U_kgDOD3e0zQ | User | false |
openclaw/openclaw | 4,272,945,354 | I_kwDOQb6kR87-r_jK | 67,488 | https://github.com/openclaw/openclaw/issues/67488 | https://api.github.com/repos/openclaw/openclaw/issues/67488 | Cron job result serialization fails on special characters in edit tool arguments | ## Bug Description
Cron job `daily-memory-review` reports `Expected ',' or '}' after property value in JSON at position 492 (line 1 column 493)` even though the agent task completes successfully.
## Reproduction
1. Set up a cron job with `sessionTarget: isolated` and `delivery.mode: announce`
2. The agent uses the `... | open | null | false | 1 | [] | [] | 2026-04-16T02:55:17Z | 2026-05-03T10:41:13Z | null | NONE | null | 20260503T120038Z | 2026-05-03T12:00:38Z | xianningqixi | 198,519,651 | U_kgDOC9UrYw | User | false |
openclaw/openclaw | 4,272,985,240 | I_kwDOQb6kR87-sJSY | 67,491 | https://github.com/openclaw/openclaw/issues/67491 | https://api.github.com/repos/openclaw/openclaw/issues/67491 | [Feature]: ：[Bug] image tool: URL images not converted to base64 before MiniMax VL API cal | ### Summary
**Environment**
- OpenClaw version: (check via `openclaw --version`)
- Node.js: v24.14.1
- OS: Windows_NT 10.0.19045
- Model: minimax/MiniMax-M2.7-highspeed
- API key type: Coding Plan (`sk-cp-*`)
**Bug Description**
When using the `image` tool with an image URL (e.g., `https://httpbin.org/image/jpeg`), ... | closed | completed | true | 1 | [
"enhancement"
] | [] | 2026-04-16T03:07:05Z | 2026-04-28T04:53:04Z | 2026-04-25T18:09:39Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | xiaopaopao361-spec | 271,329,924 | U_kgDOECwqhA | User | false |
openclaw/openclaw | 4,273,031,079 | I_kwDOQb6kR87-sUen | 67,494 | https://github.com/openclaw/openclaw/issues/67494 | https://api.github.com/repos/openclaw/openclaw/issues/67494 | Bug: Model selector not working in Chat panel (only Primary Model setting works) | # Bug: Model selector not working in Chat panel (only Primary Model setting works)
## Problem
The model selector dropdown in the **Chat panel** does not actually change the response model. Only the **Primary Model** setting in the Agent section works as expected; changing models in the Chat panel has no effect.
---
... | closed | completed | true | 1 | [] | [] | 2026-04-16T03:19:03Z | 2026-04-28T04:53:03Z | 2026-04-25T18:09:46Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | CyberRaccoonTeam | 195,688,984 | U_kgDOC6n6GA | User | false |
openclaw/openclaw | 4,273,050,294 | I_kwDOQb6kR87-sZK2 | 67,496 | https://github.com/openclaw/openclaw/issues/67496 | https://api.github.com/repos/openclaw/openclaw/issues/67496 | Slack Socket Mode: standalone @slack/socket-mode test confirms zero inbound events on v2026.4.14 | ## Summary
Slack Socket Mode connects successfully but receives zero inbound events on v2026.4.14 (and all versions since v2026.3.11). This was confirmed with a standalone test using the `@slack/socket-mode` library bundled with OpenClaw, outside of the OpenClaw gateway process.
## Standalone Reproduction
```javascr... | closed | completed | true | 2 | [] | [] | 2026-04-16T03:26:15Z | 2026-04-28T04:53:03Z | 2026-04-25T18:09:53Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | austinmao | 6,487,975 | MDQ6VXNlcjY0ODc5NzU= | User | false |
openclaw/openclaw | 4,273,069,467 | I_kwDOQb6kR87-sd2b | 67,497 | https://github.com/openclaw/openclaw/issues/67497 | https://api.github.com/repos/openclaw/openclaw/issues/67497 | [Bug]: Feishu setup/status probe leaks SDK '[info]: [ 'client ready' ]' into wizard output | ## Summary
Feishu setup/status probing leaks `@larksuiteoapi/node-sdk` internal info logs into the interactive setup wizard, so users see raw lines like:
```text
[info]: [ 'client ready' ]
```
This is not an OpenClaw user-facing message. It appears to come from the Feishu SDK client constructor and interrupts the wiz... | open | null | false | 2 | [] | [] | 2026-04-16T03:32:52Z | 2026-05-02T20:43:03Z | null | NONE | null | 20260502T210030Z | 2026-05-02T21:00:30Z | zhyongrui | 209,899,781 | U_kgDODILRBQ | User | false |
openclaw/openclaw | 4,273,112,910 | I_kwDOQb6kR87-sodO | 67,500 | https://github.com/openclaw/openclaw/issues/67500 | https://api.github.com/repos/openclaw/openclaw/issues/67500 | Main session pump fires not delivered during post-announce idle state | **Description:**
During long-run sessions with active subagents, the recurring work pump (, ) fails to deliver within the expected interval when the main session is in a post-announce idle state. The session stalls for 9–11 minutes despite the pump firing on schedule.
**Observed behavior:**
After a subagent completion... | closed | completed | true | 1 | [] | [] | 2026-04-16T03:46:46Z | 2026-04-28T04:53:02Z | 2026-04-25T18:09:59Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | A1fred-AI | 269,621,014 | U_kgDOEBIXFg | User | false |
openclaw/openclaw | 4,273,140,839 | I_kwDOQb6kR87-svRn | 67,502 | https://github.com/openclaw/openclaw/issues/67502 | https://api.github.com/repos/openclaw/openclaw/issues/67502 | [Bug]: Assistant response lost during nested lane congestion + delivery context corruption | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
## Environment
- OpenClaw: 2026.4.14 (config written by 2026.4.14, service running 2026.4.5)
- OS: Linux 6.17.0-20-generic (x64)
- Node: v24.14.1
- Channel: Discord (persistent group sessions)
- ACP Runtime: acpx + Claude C... | closed | completed | true | 0 | [
"bug",
"regression"
] | [] | 2026-04-16T03:56:09Z | 2026-04-21T04:21:24Z | 2026-04-19T02:28:57Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | Joen1750 | 5,775,449 | MDQ6VXNlcjU3NzU0NDk= | User | false |
openclaw/openclaw | 4,273,191,591 | I_kwDOQb6kR87-s7qn | 67,503 | https://github.com/openclaw/openclaw/issues/67503 | https://api.github.com/repos/openclaw/openclaw/issues/67503 | Inbound message envelope header truncates message body, causing model confusion | ## Summary
The inbound message envelope format includes the message body twice:
1. In the `System:` header line (truncated to 160 chars)
2. In full after the metadata blocks
This causes the model to sometimes reference the truncated header instead of the full message, leading to incorrect responses like "your message... | open | null | false | 2 | [] | [] | 2026-04-16T04:11:01Z | 2026-05-02T23:57:41Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | jgallowa07 | 23,087,427 | MDQ6VXNlcjIzMDg3NDI3 | User | false |
openclaw/openclaw | 4,273,239,822 | I_kwDOQb6kR87-tHcO | 67,507 | https://github.com/openclaw/openclaw/issues/67507 | https://api.github.com/repos/openclaw/openclaw/issues/67507 | CI node shard can hang after boundary+tooling combined project run | ## Summary
The `core-support-boundary` node shard can finish executing the `boundary` + `tooling` Vitest projects and then hang with no further output until the CI watchdog kills it.
## Repro
Locally, I reproduced it with:
```bash
CI=1 OPENCLAW_VITEST_NO_OUTPUT_TIMEOUT_MS=20000 node scripts/run-vitest.mjs run --confi... | closed | not_planned | true | 1 | [] | [] | 2026-04-16T04:26:55Z | 2026-04-30T04:39:53Z | 2026-04-26T05:13:00Z | CONTRIBUTOR | null | 20260430T060018Z | 2026-04-30T06:00:18Z | Pearcekieser | 5,055,971 | MDQ6VXNlcjUwNTU5NzE= | User | false |
openclaw/openclaw | 4,273,320,612 | I_kwDOQb6kR87-tbKk | 67,511 | https://github.com/openclaw/openclaw/issues/67511 | https://api.github.com/repos/openclaw/openclaw/issues/67511 | Persist per-session working context across compact/new flows | ## Summary
OpenClaw preserves semantic conversation continuity better than engineering working-context continuity. After `/compact`, `/new`, subagent runs, or temp-clone-heavy workflows, it is too easy to lose track of where work is actually happening, even when the transcript still correctly remembers what happened.
... | open | null | false | 3 | [] | [] | 2026-04-16T04:51:17Z | 2026-05-02T07:16:38Z | null | CONTRIBUTOR | null | 20260502T090030Z | 2026-05-02T09:00:30Z | yfge | 1,186,273 | MDQ6VXNlcjExODYyNzM= | User | false |
openclaw/openclaw | 4,273,386,659 | I_kwDOQb6kR87-trSj | 67,516 | https://github.com/openclaw/openclaw/issues/67516 | https://api.github.com/repos/openclaw/openclaw/issues/67516 | onboard TypeError: Cannot read properties of undefined (reading 'trim') | ## Bug
`openclaw onboard` crashes with `TypeError: Cannot read properties of undefined (reading 'trim')` at the end of the QuickStart channel selection step.
## Steps to Reproduce
1. Run `openclaw onboard` on a fresh install (or with existing config)
2. Complete security, setup mode, model/auth steps normally
3. At ... | closed | completed | true | 2 | [] | [] | 2026-04-16T05:09:13Z | 2026-04-19T04:23:27Z | 2026-04-16T06:23:56Z | NONE | null | 20260419T052035Z | 2026-04-19T05:20:35Z | lime9130461-star | 261,881,191 | U_kgDOD5v9Zw | User | false |
openclaw/openclaw | 4,273,389,059 | I_kwDOQb6kR87-tr4D | 67,517 | https://github.com/openclaw/openclaw/issues/67517 | https://api.github.com/repos/openclaw/openclaw/issues/67517 | [Bug]: openai-codex/gpt-5.4 returns Cloudflare HTML and gets misclassified as rate_limit / DNS, leaving TUI stuck | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
openai-codex/gpt-5.4 stalls indefinitely in the TUI. Logs show Cloudflare / HTML responses, but OpenClaw misclassifies the failure as API rate limit reached, DNS lookup for provider endpoint failed, or raw <html>....
### S... | closed | completed | true | 3 | [
"bug",
"regression"
] | [] | 2026-04-16T05:10:01Z | 2026-04-19T04:23:15Z | 2026-04-16T12:49:55Z | NONE | null | 20260419T052035Z | 2026-04-19T05:20:35Z | 1jsamuel | 114,688,273 | U_kgDOBtYBEQ | User | false |
openclaw/openclaw | 4,273,393,671 | I_kwDOQb6kR87-ttAH | 67,519 | https://github.com/openclaw/openclaw/issues/67519 | https://api.github.com/repos/openclaw/openclaw/issues/67519 | [GPT 5.4 v3 Phase 3.B] First-class plan rendering across CLI, Control UI, and messaging channels | ## GPT 5.4 Enhancement v3 — Phase 3.B
**Tracking: #66345**
**Priority: P0 — Highest user-visible impact per §7.4 ranking**
**Depends on: #67514 (task-system parity — cancelled status + activeForm field)**
## Problem
Today `update_plan` emits structured `AgentPlanEventData` events, but most channel plugins show the ra... | closed | completed | true | 1 | [] | [] | 2026-04-16T05:11:27Z | 2026-04-25T04:17:53Z | 2026-04-22T10:35:06Z | CONTRIBUTOR | null | 20260425T060021Z | 2026-04-25T06:00:21Z | 100yenadmin | 239,388,517 | U_kgDODkTHZQ | User | false |
openclaw/openclaw | 4,273,395,659 | I_kwDOQb6kR87-ttfL | 67,520 | https://github.com/openclaw/openclaw/issues/67520 | https://api.github.com/repos/openclaw/openclaw/issues/67520 | [GPT 5.4 v3 Phase 3.C] Plan-mode entry points + approval UX (builds on #61845) | ## GPT 5.4 Enhancement v3 — Phase 3.C
**Tracking: #66345**
**Priority: P1 — The marquee Claude Code-quality feature (opt-in only)**
**Hard depends on: #61845 merging (plan-mode runtime from nick0809cn-ux)**
**Also depends on: #67514 (task-system parity), #67519 (plan rendering)**
## Context
PR #61845 already builds t... | closed | completed | true | 1 | [] | [] | 2026-04-16T05:12:04Z | 2026-04-25T04:17:52Z | 2026-04-22T10:35:09Z | CONTRIBUTOR | null | 20260425T060021Z | 2026-04-25T06:00:21Z | 100yenadmin | 239,388,517 | U_kgDODkTHZQ | User | false |
openclaw/openclaw | 4,273,397,179 | I_kwDOQb6kR87-tt27 | 67,521 | https://github.com/openclaw/openclaw/issues/67521 | https://api.github.com/repos/openclaw/openclaw/issues/67521 | [GPT 5.4 v3 Phase 3.D] qa-lab parity scenarios + verification harness updates | ## GPT 5.4 Enhancement v3 — Phase 3.D
**Tracking: #66345**
**Priority: P1 — Closes the verification loop on the entire parity sprint**
**Depends on: #67512, #67518, #67514 (final sprint PRs) + #67519, #67520 (Phase 3 PRs)**
**Builds on: #64441 (first-wave parity harness), #65664 (parity proof slice)**
## Problem
The ... | closed | completed | true | 1 | [] | [] | 2026-04-16T05:12:28Z | 2026-04-25T04:17:51Z | 2026-04-22T10:35:12Z | CONTRIBUTOR | null | 20260425T060021Z | 2026-04-25T06:00:21Z | 100yenadmin | 239,388,517 | U_kgDODkTHZQ | User | false |
openclaw/openclaw | 4,273,398,479 | I_kwDOQb6kR87-tuLP | 67,522 | https://github.com/openclaw/openclaw/issues/67522 | https://api.github.com/repos/openclaw/openclaw/issues/67522 | [GPT 5.4 Phase 4.1] Skill-driven plan templates | ## GPT 5.4 Enhancement — Phase 4.1
**Tracking: #66345**
**Priority: P2 — Power-user differentiator (not parity-critical)**
**Depends on: #67514 (task-system parity), #67520 (plan mode)**
**Confidence: 75% (novel feature — not in Hermes or Claude Code)**
## Concept
A skill (`.openclaw/skills/<name>/SKILL.md`) can decl... | closed | completed | true | 1 | [] | [] | 2026-04-16T05:12:48Z | 2026-04-25T04:17:51Z | 2026-04-22T10:35:14Z | CONTRIBUTOR | null | 20260425T060021Z | 2026-04-25T06:00:21Z | 100yenadmin | 239,388,517 | U_kgDODkTHZQ | User | false |
openclaw/openclaw | 4,273,399,804 | I_kwDOQb6kR87-tuf8 | 67,523 | https://github.com/openclaw/openclaw/issues/67523 | https://api.github.com/repos/openclaw/openclaw/issues/67523 | [GPT 5.4 Phase 4.2] Cross-session task coordination (Tasks API style) | ## GPT 5.4 Enhancement — Phase 4.2
**Tracking: #66345**
**Priority: P2 — Power-user feature for multi-agent workflows**
**Depends on: #67514 (task-system parity), #67520 (plan mode)**
**Confidence: 70% (complex; Claude Code Tasks API does this but Hermes does not)**
## Concept
Multiple sessions / multiple agents can ... | closed | completed | true | 1 | [] | [] | 2026-04-16T05:13:12Z | 2026-04-25T04:17:50Z | 2026-04-22T10:35:17Z | CONTRIBUTOR | null | 20260425T060021Z | 2026-04-25T06:00:21Z | 100yenadmin | 239,388,517 | U_kgDODkTHZQ | User | false |
openclaw/openclaw | 4,273,415,363 | I_kwDOQb6kR87-tyTD | 67,524 | https://github.com/openclaw/openclaw/issues/67524 | https://api.github.com/repos/openclaw/openclaw/issues/67524 | [Bug]: trusted-proxy validator rejects `auth.token` even though #54536 requires it for local-direct fallback | ## Summary
PR #54536 (merged Mar 29, shipped from v2026.4.1 onwards) added a local-direct fallback in `authorizeGatewayConnect` so that requests arriving on loopback without proxy identity headers authenticate via `auth.token` instead of being rejected. This is the intended fix for the long-standing issue that interna... | closed | not_planned | true | 3 | [] | [] | 2026-04-16T05:16:54Z | 2026-04-30T04:39:52Z | 2026-04-26T05:13:04Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | GanemoCorp | 46,256,429 | MDQ6VXNlcjQ2MjU2NDI5 | User | false |
openclaw/openclaw | 4,273,420,638 | I_kwDOQb6kR87-tzle | 67,525 | https://github.com/openclaw/openclaw/issues/67525 | https://api.github.com/repos/openclaw/openclaw/issues/67525 | [Bug]: openclaw doctor crashes in 2026.4.14: clearPluginInteractiveHandlersState calls .clear() on callback dedupe state | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
## Summary
`openclaw doctor` crashes on OpenClaw 2026.4.14 with:
TypeError: Cannot read properties of undefined (reading 'clear')
This happens during the "Memory search" section, but the actual root cause appears to be in ... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-16T05:18:09Z | 2026-04-28T04:52:57Z | 2026-04-25T18:11:14Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | luoshuzheng-bytedance | 256,737,957 | U_kgDOD02CpQ | User | false |
openclaw/openclaw | 4,273,453,692 | I_kwDOQb6kR87-t7p8 | 67,527 | https://github.com/openclaw/openclaw/issues/67527 | https://api.github.com/repos/openclaw/openclaw/issues/67527 | WebChat: async command completion notifications incorrectly shown in user message bubbles | ## Bug type
UI/Display bug
## Summary
In WebChat/Control UI, async command completion notifications (e.g., `System (untrusted): [2026-04-16 13:19:39 GMT+8] Exec completed...`) are incorrectly displayed **inside user message bubbles** instead of being shown as system messages or hidden entirely.
## Environment
- Open... | closed | completed | true | 2 | [] | [] | 2026-04-16T05:26:24Z | 2026-04-28T04:52:56Z | 2026-04-25T18:11:21Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | ChengZhuo2025 | 203,413,187 | U_kgDODB_Www | User | false |
openclaw/openclaw | 4,273,476,882 | I_kwDOQb6kR87-uBUS | 67,529 | https://github.com/openclaw/openclaw/issues/67529 | https://api.github.com/repos/openclaw/openclaw/issues/67529 | feat(channel): VK (VKontakte) community channel plugin | ## Summary
I've built and published a full **VK (VKontakte) channel plugin** for OpenClaw — the first one for Russia's largest social network (100M+ monthly active users).
**Repo:** https://github.com/fibbersha-hub/openclaw-vk-plugin
## Features
- **Messaging**: Send/receive/edit/delete messages, inline & standard ... | closed | not_planned | true | 2 | [] | [] | 2026-04-16T05:32:41Z | 2026-04-28T04:52:55Z | 2026-04-25T18:11:28Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | fibbersha-hub | 246,649,733 | U_kgDODrOThQ | User | false |
openclaw/openclaw | 4,273,478,034 | I_kwDOQb6kR87-uBmS | 67,530 | https://github.com/openclaw/openclaw/issues/67530 | https://api.github.com/repos/openclaw/openclaw/issues/67530 | UI: Context window percentage meter always shows 0% for Anthropic/Claude models | ## Bug Report
### Description
The context window usage meter in the Control UI always displays **0%** when using Anthropic/Claude models, even though the actual usage is tracked and reported correctly elsewhere.
### Steps to Reproduce
1. Configure OpenClaw with an Anthropic Claude model (e.g. `anthropic/claude-sonn... | closed | completed | true | 1 | [] | [] | 2026-04-16T05:33:00Z | 2026-04-28T04:52:54Z | 2026-04-25T18:11:35Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Clarkar00 | 133,609,856 | U_kgDOB_a5gA | User | false |
openclaw/openclaw | 4,273,497,864 | I_kwDOQb6kR87-uGcI | 67,532 | https://github.com/openclaw/openclaw/issues/67532 | https://api.github.com/repos/openclaw/openclaw/issues/67532 | Z.AI Coding Plan baseUrl gets spurious `/v1` appended in agent-scoped `models.json` → 404 → LiveSessionModelSwitchError forces switch to `openai/gpt-5.4` | # Z.AI Coding Plan baseUrl gets spurious `/v1` appended in agent-scoped `models.json` → 404 → LiveSessionModelSwitchError forces switch to `openai/gpt-5.4`
## Summary
When a Z.AI provider is configured with the **GLM Coding Plan Global** endpoint (`https://api.z.ai/api/coding/paas/v4`) matching the canonical constant... | closed | completed | true | 3 | [] | [] | 2026-04-16T05:38:52Z | 2026-04-30T04:36:44Z | 2026-04-27T05:22:56Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | tuxclaw | 261,464,027 | U_kgDOD5Wf2w | User | false |
openclaw/openclaw | 4,273,521,833 | I_kwDOQb6kR87-uMSp | 67,537 | https://github.com/openclaw/openclaw/issues/67537 | https://api.github.com/repos/openclaw/openclaw/issues/67537 | [Feature]: GitHub 登录持续失败，手机和电脑都试过了，Sign in failed. Please try again. | ### Summary
<img width="402" height="75" alt="Image" src="https://github.com/user-attachments/assets/d746ee45-6e41-436d-b430-8e89b6cddc28" />
### Problem to solve
登录持续失败，手机和电脑都试过了，Sign in failed. Please try again.
### Proposed solution
登录持续失败，手机和电脑都试过了，Sign in failed. Please try again.
### Alternatives considered... | closed | not_planned | true | 1 | [
"enhancement"
] | [] | 2026-04-16T05:46:22Z | 2026-04-30T04:39:51Z | 2026-04-26T05:13:08Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | jinkang19940922 | 96,525,444 | U_kgDOBcDchA | User | false |
openclaw/openclaw | 4,273,523,902 | I_kwDOQb6kR87-uMy- | 67,539 | https://github.com/openclaw/openclaw/issues/67539 | https://api.github.com/repos/openclaw/openclaw/issues/67539 | [Feature]: Add provider-specific TTS prompt hints | ## Summary
Add provider-specific, model-aware TTS prompt hints so agents know which expressive controls are valid for the active speech provider.
## Problem to solve
OpenClaw already has a generic TTS runtime hint that tells agents they can use `[[tts:...]]` directives and optional `[[tts:text]]...[[/tts:text]]` blo... | open | null | false | 2 | [] | [] | 2026-04-16T05:46:58Z | 2026-05-02T23:57:42Z | null | CONTRIBUTOR | null | 20260503T000641Z | 2026-05-03T00:06:41Z | barronlroth | 5,567,884 | MDQ6VXNlcjU1Njc4ODQ= | User | false |
openclaw/openclaw | 4,273,547,320 | I_kwDOQb6kR87-uSg4 | 67,543 | https://github.com/openclaw/openclaw/issues/67543 | https://api.github.com/repos/openclaw/openclaw/issues/67543 | Add k2.6-code-preview model support for Kimi/Moonshot providers | ## Request: Add k2.6-code-preview model support
### Context
NousResearch/hermes-agent recently landed support for the k2.6-code-preview model across their Kimi providers (kimi-coding, kimi-coding-cn, moonshot). See PR [#10751](https://github.com/NousResearch/hermes-agent/pull/10751).
We would like to request similar ... | closed | completed | true | 3 | [] | [] | 2026-04-16T05:53:35Z | 2026-04-23T04:22:12Z | 2026-04-21T01:28:10Z | NONE | null | 20260423T060020Z | 2026-04-23T06:00:20Z | nicoechaniz | 198,112 | MDQ6VXNlcjE5ODExMg== | User | false |
openclaw/openclaw | 4,273,554,799 | I_kwDOQb6kR87-uUVv | 67,544 | https://github.com/openclaw/openclaw/issues/67544 | https://api.github.com/repos/openclaw/openclaw/issues/67544 | Gateway restart can fail on Windows with EADDRINUSE after crash or overlapping startup paths | ## Summary
On Windows, the OpenClaw gateway can fail to restart cleanly after a crash, abnormal exit, or overlapping startup attempt because port `18789` remains held by an existing/stale Node process. Subsequent restart attempts fail with `EADDRINUSE`, leaving the UI unavailable until the lingering process is manuall... | closed | completed | true | 1 | [] | [] | 2026-04-16T05:55:19Z | 2026-04-28T04:52:54Z | 2026-04-25T18:11:49Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | MichealBreedlove | 8,867,704 | MDQ6VXNlcjg4Njc3MDQ= | User | false |
openclaw/openclaw | 4,273,579,519 | I_kwDOQb6kR87-uaX_ | 67,546 | https://github.com/openclaw/openclaw/issues/67546 | https://api.github.com/repos/openclaw/openclaw/issues/67546 | [Bug]: Onboard Feature Bug | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
```
◇ Select channel (QuickStart)
│ Skip for now
TypeError: Cannot read properties of undefined (reading 'trim')
ubuntu@ip-172-31-20-200:~$
```
After selecting skip for now when asked to select a way to co... | closed | completed | true | 2 | [
"bug",
"bug:behavior"
] | [] | 2026-04-16T06:01:21Z | 2026-04-19T04:23:27Z | 2026-04-16T06:23:56Z | NONE | null | 20260419T052035Z | 2026-04-19T05:20:35Z | WinstonJunCong | 54,414,639 | MDQ6VXNlcjU0NDE0NjM5 | User | false |
openclaw/openclaw | 4,273,620,453 | I_kwDOQb6kR87-ukXl | 67,547 | https://github.com/openclaw/openclaw/issues/67547 | https://api.github.com/repos/openclaw/openclaw/issues/67547 | [Bug]: A2UI bundle/hash depends on ui/node_modules presence | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Running `pnpm build` on the same checkout rewrites `src/canvas-host/a2ui/.bundle.hash` and `src/canvas-host/a2ui/a2ui.bundle.js` depending on whether `ui/node_modules` exists, so the committed A2UI prebuilt art... | open | null | false | 2 | [] | [] | 2026-04-16T06:12:46Z | 2026-05-03T08:39:30Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | Tony-ooo | 45,898,990 | MDQ6VXNlcjQ1ODk4OTkw | User | false |
openclaw/openclaw | 4,273,620,520 | I_kwDOQb6kR87-ukYo | 67,548 | https://github.com/openclaw/openclaw/issues/67548 | https://api.github.com/repos/openclaw/openclaw/issues/67548 | Caracteres chinos o texto extraño se inserta en respuestas durante conversaciones largas | ## Bug Description
In long conversations where the `exec` command is used multiple times, Chinese characters (for example: 这些, 反复, 不确定) and text in other languages appear inserted into the assistant’s responses without being requested by the user.
## Steps to Reproduce
1. Maintain a long conversation
2. Use the `exe... | closed | completed | true | 1 | [] | [] | 2026-04-16T06:12:47Z | 2026-04-28T04:52:53Z | 2026-04-25T18:11:56Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Tonisg91 | 56,615,093 | MDQ6VXNlcjU2NjE1MDkz | User | false |
openclaw/openclaw | 4,273,631,912 | I_kwDOQb6kR87-unKo | 67,549 | https://github.com/openclaw/openclaw/issues/67549 | https://api.github.com/repos/openclaw/openclaw/issues/67549 | [Bug] memory_search tool throws 'Cannot read properties of undefined' in OpenClaw 4.14 (CLI works fine) | ## Description
In OpenClaw 4.14 (323493f), the `memory_search` agent tool call fails with error:
```
Cannot read properties of undefined (reading 'get')
```
However, the `openclaw memory search` CLI command works perfectly fine with the same query.
## Environment
- Version: OpenClaw 2026.4.14 (323493f)
- OS: Linux ... | closed | completed | true | 1 | [] | [] | 2026-04-16T06:15:54Z | 2026-04-28T04:52:52Z | 2026-04-25T18:12:03Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | ztx1447321467 | 101,324,257 | U_kgDOBgoV4Q | User | false |
openclaw/openclaw | 4,273,635,563 | I_kwDOQb6kR87-uoDr | 67,552 | https://github.com/openclaw/openclaw/issues/67552 | https://api.github.com/repos/openclaw/openclaw/issues/67552 | [Feature]: RFC: Tool Response Freshness Metadata | ### Summary
Standardize tool response metadata to expose data freshness information (`cached`, `age_ms`, `ttl_ms`), enabling agents to make informed decisions about whether cached data is acceptable for their current task.
### Problem to solve
Many OpenClaw tools return data that may be cached, stale, or outdated (e... | open | null | false | 1 | [
"enhancement"
] | [] | 2026-04-16T06:16:54Z | 2026-05-03T08:25:27Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | CaptianZoro | 266,676,182 | U_kgDOD-Un1g | User | false |
openclaw/openclaw | 4,273,642,019 | I_kwDOQb6kR87-upoj | 67,553 | https://github.com/openclaw/openclaw/issues/67553 | https://api.github.com/repos/openclaw/openclaw/issues/67553 | [Feature Request] Dynamic Heartbeat Checklist: Parse HEARTBEAT.md to Determine Active Checks | ## Description
Currently, the heartbeat logic in OpenClaw includes a **static, hardcoded checklist** of checks (e.g., Calendar, Email, Reminders, Agent Refresh Schedule) regardless of whether those sections exist in the user's `HEARTBEAT.md` file.
Even after a user removes a section from `HEARTBEAT.md` and restarts th... | closed | completed | true | 1 | [] | [] | 2026-04-16T06:18:46Z | 2026-04-30T04:39:50Z | 2026-04-26T05:15:15Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | pabbas29 | 61,113,904 | MDQ6VXNlcjYxMTEzOTA0 | User | false |
openclaw/openclaw | 4,273,735,888 | I_kwDOQb6kR87-vAjQ | 67,562 | https://github.com/openclaw/openclaw/issues/67562 | https://api.github.com/repos/openclaw/openclaw/issues/67562 | memory-core: dreaming narrative push fails on Feishu due to hardcoded "default" target | ### Bug Description
When `memory-core.dreaming.enabled` is true and using Feishu channel, the dreaming plugin attempts to push narrative diary entries with a hardcoded `"default"` target. This causes Feishu API to reject the request because Feishu requires a specific `open_id` (e.g., `ou_xxx`) instead of the generic `"... | closed | completed | true | 1 | [] | [] | 2026-04-16T06:42:54Z | 2026-04-28T04:52:51Z | 2026-04-25T18:12:18Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | JiayangSUNwinter | 126,450,746 | U_kgDOB4l8Og | User | false |
openclaw/openclaw | 4,273,753,578 | I_kwDOQb6kR87-vE3q | 67,563 | https://github.com/openclaw/openclaw/issues/67563 | https://api.github.com/repos/openclaw/openclaw/issues/67563 | Feature Request: Streaming Tool Execution Progress Rendering | ## Feature Request: 流式工具执行进度渲染
### 问题描述
目前 agent 执行任务时，所有工具调用是一次性完成并一次性返回结果。用户无法实时看到 agent 每一步在做什么。
### 期望效果（参考 Hermes Agent）
Hermes Agent 框架在执行任务时，会在每个工具调用完成后立即推送一个"进度块"给用户：
```
收到任务：帮我搜一下今天的科技新闻
↓
[1/3] 正在搜索... ●○○
[1/3] 搜索完成 ●●○
[2/3] 正在分析... ●●○
[2/3] 分析完成 ●●●
[3/3] 正在整理... ●●●
↓
最终结果
```
### 关键特点
1. **实时呈现*... | closed | completed | true | 1 | [] | [] | 2026-04-16T06:46:30Z | 2026-04-28T04:52:51Z | 2026-04-25T18:12:25Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | qiukui666 | 169,086,124 | U_kgDOChQMrA | User | false |
openclaw/openclaw | 4,273,760,800 | I_kwDOQb6kR87-vGog | 67,564 | https://github.com/openclaw/openclaw/issues/67564 | https://api.github.com/repos/openclaw/openclaw/issues/67564 | [Bug]: openclaw-weixin getUpdates long-poll intermittently fails with UND_ERR_HEADERS_TIMEOUT and enters 30s backoff loop | ## Description
### Bug type
Behavior bug (incorrect output/state without crash)
### Summary
`@tencent-weixin/openclaw-weixin` intermittently fails during the long-poll `getUpdates` loop with `TypeError: fetch failed`. After 3 consecutive failures it enters the built-in 30s backoff, which makes the Weixin channel app... | closed | not_planned | true | 2 | [] | [] | 2026-04-16T06:47:59Z | 2026-04-28T04:52:50Z | 2026-04-25T18:12:32Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Theo-jobs | 177,595,285 | U_kgDOCpXjlQ | User | false |
openclaw/openclaw | 4,273,819,088 | I_kwDOQb6kR87-vU3Q | 67,565 | https://github.com/openclaw/openclaw/issues/67565 | https://api.github.com/repos/openclaw/openclaw/issues/67565 | [Feature]: Add --daemon-env flag to openclaw onboard --install-daemon | ## Summary
Add a repeatable `--daemon-env KEY=VALUE` flag to `openclaw onboard --install-daemon` so that extra `Environment=` lines can be injected into the generated systemd unit at bootstrap time.
## Problem to solve
When deploying a gateway with `networking.mode: docker` (or `headscale`, or any LAN-bound configur... | closed | not_planned | true | 1 | [] | [] | 2026-04-16T06:59:37Z | 2026-04-30T04:39:49Z | 2026-04-26T05:15:20Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | dlebee | 36,390,643 | MDQ6VXNlcjM2MzkwNjQz | User | false |
openclaw/openclaw | 4,273,898,961 | I_kwDOQb6kR87-voXR | 67,566 | https://github.com/openclaw/openclaw/issues/67566 | https://api.github.com/repos/openclaw/openclaw/issues/67566 | [Bug]: memorySearch tool fails with 'Cannot read properties of undefined (reading get)' - memory-core plugin meta table empty on v2026.4.14 | ## Description
The `memory_search` tool consistently fails with:
```
Cannot read properties of undefined (reading 'get')
Memory search is unavailable due to an embedding/provider error.
```
This occurs even though:
- SiliconFlow embedding API is working (curl confirms 1024-dim vectors returned)
- SQLite index file e... | closed | completed | true | 2 | [] | [] | 2026-04-16T07:17:48Z | 2026-04-30T04:39:48Z | 2026-04-26T05:15:25Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | Bowerai | 17,424,365 | MDQ6VXNlcjE3NDI0MzY1 | User | false |
openclaw/openclaw | 4,273,920,079 | I_kwDOQb6kR87-vthP | 67,568 | https://github.com/openclaw/openclaw/issues/67568 | https://api.github.com/repos/openclaw/openclaw/issues/67568 | Session memory files use UTC timestamps, causing date mismatch with local timezone | ## Bug Description
Session memory files (e.g., `.reset.2026-04-16T04-58-33.868Z`) use UTC timestamps in their filenames, but the user is located in `Asia/Shanghai (GMT+8)`.
When the memory/heartbeat system derives dates from these timestamps, a session that starts at `2026-04-16 04:00 UTC` (which is `2026-04-16 12:00... | closed | completed | true | 1 | [] | [] | 2026-04-16T07:22:27Z | 2026-04-30T04:36:43Z | 2026-04-27T05:23:02Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | xunx33 | 269,385,656 | U_kgDOEA5_uA | User | false |
openclaw/openclaw | 4,274,022,793 | I_kwDOQb6kR87-wGmJ | 67,573 | https://github.com/openclaw/openclaw/issues/67573 | https://api.github.com/repos/openclaw/openclaw/issues/67573 | /new and /reset commands should clear session-level model override to respect agents.defaults.model configuration | ### Problem Description
When a user explicitly changes the model via `/model <model-id>`, the session-level model override persists even after running `/new` or `/reset`. This causes the session to continue using the previously selected model instead of falling back to the configured default (`agents.defaults.model.pr... | closed | completed | false | 1 | [
"close:duplicate",
"dedupe:child"
] | [] | 2026-04-16T07:44:54Z | 2026-04-25T08:29:35Z | 2026-04-25T08:29:35Z | NONE | null | 20260425T090015Z | 2026-04-25T09:00:15Z | MiaoDX | 8,840,792 | MDQ6VXNlcjg4NDA3OTI= | User | false |
openclaw/openclaw | 4,274,026,716 | I_kwDOQb6kR87-wHjc | 67,574 | https://github.com/openclaw/openclaw/issues/67574 | https://api.github.com/repos/openclaw/openclaw/issues/67574 | [Bug]: An Error aftet set channel when use install dameon | ### Bug type
Crash (process/app exits or hangs)
### Beta release blocker
No
### Summary
In the version of **2026.4.14**, when use `openclaw onboard --install-daemon` to do setup, after set the channel, then terminal gave an Error: `**TypeError: Cannot read properties of undefined (reading 'trim')**`
Claude Code w... | closed | completed | true | 1 | [
"bug",
"bug:crash"
] | [] | 2026-04-16T07:45:46Z | 2026-04-23T04:22:39Z | 2026-04-20T06:27:44Z | NONE | null | 20260423T060020Z | 2026-04-23T06:00:20Z | wanshubin | 67,248,207 | MDQ6VXNlcjY3MjQ4MjA3 | User | false |
openclaw/openclaw | 4,274,042,817 | I_kwDOQb6kR87-wLfB | 67,575 | https://github.com/openclaw/openclaw/issues/67575 | https://api.github.com/repos/openclaw/openclaw/issues/67575 | [Bug]: OpenRouter responses received but not returned to user (payloads=0, incomplete turn) | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
OpenClaw successfully sends requests to OpenRouter, but responses are not surfaced to the user across all interfaces (Telegram, CLI, TUI).
### Steps to reproduce
Configure OpenRouter model
Restart gateway
Run... | closed | completed | true | 6 | [
"bug",
"bug:behavior"
] | [] | 2026-04-16T07:48:55Z | 2026-04-21T04:22:02Z | 2026-04-18T16:02:13Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | Deterrence89 | 128,595,413 | U_kgDOB6o11Q | User | false |
openclaw/openclaw | 4,274,072,175 | I_kwDOQb6kR87-wSpv | 67,576 | https://github.com/openclaw/openclaw/issues/67576 | https://api.github.com/repos/openclaw/openclaw/issues/67576 | [Bug]: infer model run ignores OpenRouter provider in model string and falls back to openai-codex | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
In our current setup, normal OpenRouter-backed web calls appear in OpenRouter logs, so OpenRouter connectivity itself is working.
However, when running infer model run with an OpenRouter-prefixed model string, for example:
... | closed | completed | true | 2 | [
"bug",
"regression"
] | [] | 2026-04-16T07:54:30Z | 2026-04-21T04:21:51Z | 2026-04-18T17:13:09Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | AtlasAI2026 | 275,035,787 | U_kgDOEGS2iw | User | false |
openclaw/openclaw | 4,274,075,946 | I_kwDOQb6kR87-wTkq | 67,578 | https://github.com/openclaw/openclaw/issues/67578 | https://api.github.com/repos/openclaw/openclaw/issues/67578 | @homebridge/ciao assertion failure causes complete gateway crash on malformed mDNS packet from macOS peer | ## Summary
The OpenClaw gateway crashes with an uncaught `AssertionError` when processing mDNS/Bonjour packets from a peer macOS device. The crash appears to be triggered by inconsistent IP address/netmask versions in mDNS advertisements, possibly related to dual-stack IPv4/IPv6 handling or VNC screen sharing traffic.
... | closed | completed | true | 4 | [] | [] | 2026-04-16T07:55:14Z | 2026-04-29T04:38:20Z | 2026-04-26T09:47:51Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | zenassist26-create | 261,834,431 | U_kgDOD5tGvw | User | false |
openclaw/openclaw | 4,274,100,932 | I_kwDOQb6kR87-wZrE | 67,580 | https://github.com/openclaw/openclaw/issues/67580 | https://api.github.com/repos/openclaw/openclaw/issues/67580 | Dreaming promotion writes raw Candidate data into MEMORY.md (no distillation) | ## Summary
Dreaming short-term promotion writes raw `Candidate: User/Assistant: ... confidence: 0.00 evidence: ...` lines directly into MEMORY.md without any summarization or distillation.
## Problem
The promotion pipeline works as follows:
1. `active-memory` extracts recall entries from conversations → stored in `... | open | null | false | 4 | [] | [] | 2026-04-16T08:00:14Z | 2026-05-02T12:53:58Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | fkc0p | 219,618,780 | U_kgDODRcd3A | User | false |
openclaw/openclaw | 4,274,107,042 | I_kwDOQb6kR87-wbKi | 67,582 | https://github.com/openclaw/openclaw/issues/67582 | https://api.github.com/repos/openclaw/openclaw/issues/67582 | Webchat: User messages not visible immediately after sending (optimistic rendering missing) | When sending a message in the WebChat UI, the text I typed disappears/hides after hitting submit. I can see the assistant's response eventually, but my own sent message doesn't appear in the chat history right away — there's a noticeable delay (several seconds) before it renders.
This creates a poor user experience:
-... | closed | completed | true | 2 | [] | [] | 2026-04-16T08:01:25Z | 2026-04-28T04:52:49Z | 2026-04-25T18:12:40Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | wangjinfeng | 5,388,846 | MDQ6VXNlcjUzODg4NDY= | User | false |
openclaw/openclaw | 4,274,140,328 | I_kwDOQb6kR87-wjSo | 67,585 | https://github.com/openclaw/openclaw/issues/67585 | https://api.github.com/repos/openclaw/openclaw/issues/67585 | [Feature]: The earliest batch of openclaw-Agent technical exchange whatsapp group | ### Summary
<img width="1260" height="2844" alt="Image" src="https://github.com/user-attachments/assets/f4492255-c39e-4436-83d3-15d838b9ed3d" />
### Problem to solve
<img width="1260" height="2844" alt="Image" src="https://github.com/user-attachments/assets/b97b8aec-f829-4f49-9d07-bac8de66df45" />
### Proposed solu... | closed | not_planned | true | 1 | [
"enhancement"
] | [] | 2026-04-16T08:08:29Z | 2026-04-28T04:52:48Z | 2026-04-25T18:12:47Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | comqx | 30,148,386 | MDQ6VXNlcjMwMTQ4Mzg2 | User | false |
openclaw/openclaw | 4,274,166,014 | I_kwDOQb6kR87-wpj- | 67,587 | https://github.com/openclaw/openclaw/issues/67587 | https://api.github.com/repos/openclaw/openclaw/issues/67587 | Discord outbound messages should reliably convert known @AgentName references into real <@userId> mentions | I’m running a multi-agent setup on Discord where each OpenClaw agent has its own bot/user account.
I ran into a practical problem: when one agent tries to notify another by writing something like `@Vladislava`, Discord does not always treat that as a real mention. In the failing case, it gets sent as plain text instea... | closed | completed | true | 2 | [] | [] | 2026-04-16T08:13:53Z | 2026-05-04T04:34:23Z | 2026-05-02T02:54:24Z | NONE | null | 20260504T060031Z | 2026-05-04T06:00:31Z | McoreD | 48,652 | MDQ6VXNlcjQ4NjUy | User | false |
openclaw/openclaw | 4,274,299,364 | I_kwDOQb6kR87-xKHk | 67,590 | https://github.com/openclaw/openclaw/issues/67590 | https://api.github.com/repos/openclaw/openclaw/issues/67590 | [Feature] 渠道绑定配置 - 工具调用过程流式输出到频道 | ## 功能描述
希望支持**按渠道绑定配置工具调用过程的流式输出**，让用户可以在飞书/微信/Discord 等频道中实时查看 AI 的工具调用记录和内部执行过程，类似 Web Control UI 的完整信息展示。
## 使用场景
### 场景 1：长任务监督
- 用户启动一个需要多次工具调用的复杂任务（如市场调研、文献综述）
- 任务执行 10-30 分钟，期间频道中无任何反馈
- 用户无法判断 AI 是否在工作、进度如何、是否卡住
### 场景 2：调试和审计
- 开发者需要查看 AI 实际调用了哪些工具、参数是什么
- 质量审核需要追溯 AI 的决策过程和信息来源
- 当前只能查看 Web UI 或 session ... | closed | not_planned | true | 2 | [] | [] | 2026-04-16T08:36:43Z | 2026-04-28T04:52:48Z | 2026-04-25T18:13:02Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | jinzhu1991 | 199,324,600 | U_kgDOC-FzuA | User | false |
openclaw/openclaw | 4,274,376,351 | I_kwDOQb6kR87-xc6f | 67,593 | https://github.com/openclaw/openclaw/issues/67593 | https://api.github.com/repos/openclaw/openclaw/issues/67593 | feat: add Kimi/Moonshot provider usage and balance display | ## Summary
The kimi-coding provider plugin does not expose usage/balance information in the OpenClaw app panel, `/quota` command, or `openclaw status` surfaces. Other providers like zai, anthropic, and openai-codex already show balance and usage metrics.
## Current behavior
- The `usageProviders` list in `src/infra/... | open | null | false | 1 | [] | [] | 2026-04-16T08:47:46Z | 2026-05-03T07:44:46Z | null | CONTRIBUTOR | null | 20260503T090142Z | 2026-05-03T09:01:42Z | dario-github | 40,749,119 | MDQ6VXNlcjQwNzQ5MTE5 | User | false |
openclaw/openclaw | 4,274,386,369 | I_kwDOQb6kR87-xfXB | 67,594 | https://github.com/openclaw/openclaw/issues/67594 | https://api.github.com/repos/openclaw/openclaw/issues/67594 | Temporal decay should apply to dated files in memory/ subdirectories | ## Problem
Current `DATED_MEMORY_PATH_RE` regex only matches `memory/YYYY-MM-DD.md`, not dated files in subdirectories like `memory/daily-log/2026-04-10.md`.
Archived dated files are incorrectly treated as evergreen, causing them to rank higher than recent logs.
---
## Proposed solution
**Option A: Fix the regex (... | open | null | false | 2 | [] | [] | 2026-04-16T08:49:11Z | 2026-05-04T18:13:30Z | null | NONE | null | 20260504T210031Z | 2026-05-04T21:00:31Z | sunvember | 8,868,842 | MDQ6VXNlcjg4Njg4NDI= | User | false |
openclaw/openclaw | 4,274,421,080 | I_kwDOQb6kR87-xn1Y | 67,595 | https://github.com/openclaw/openclaw/issues/67595 | https://api.github.com/repos/openclaw/openclaw/issues/67595 | openclaw gateway install --force still embeds OPENCLAW_GATEWAY_TOKEN in LaunchAgent plist; openclaw channels list throws TypeError: fetch failed with no channels configured | ## Version
OpenClaw 2026.4.14 (323493f)
## Environment
- macOS 26.4.1 (arm64)
- node 25.9.0
- Gateway mode: local
- LaunchAgent-managed gateway
- Gateway bind: loopback
- Gateway port: 18789
## Issue 1
`openclaw gateway install --force` reports success, but the generated LaunchAgent plist still embeds `OPENCLAW_GA... | open | null | false | 3 | [] | [] | 2026-04-16T08:54:38Z | 2026-05-03T07:44:49Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | chouti | 27,688 | MDQ6VXNlcjI3Njg4 | User | false |
openclaw/openclaw | 4,274,448,333 | I_kwDOQb6kR87-xufN | 67,597 | https://github.com/openclaw/openclaw/issues/67597 | https://api.github.com/repos/openclaw/openclaw/issues/67597 | [Bug]: TypeError: Cannot read properties of undefined (reading 'trim') | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
升级到2026.4.14之后，进行openclaw onboard的时候，来到Select channel (QuickStart)之后就会报错，错误提示：TypeError: Cannot read properties of undefined (reading 'trim')
已回退到2026.4.11，问题消失，确认是版本问题。
openclaw doctor 也无法解决此问题。
<img width="616" height="26... | closed | completed | true | 2 | [
"bug",
"close:already-fixed",
"dedupe:child",
"regression"
] | [
"vincentkoc"
] | 2026-04-16T08:59:32Z | 2026-04-28T05:28:56Z | 2026-04-25T09:38:02Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | s1sunny | 51,485,377 | MDQ6VXNlcjUxNDg1Mzc3 | User | false |
openclaw/openclaw | 4,274,531,277 | I_kwDOQb6kR87-yCvN | 67,598 | https://github.com/openclaw/openclaw/issues/67598 | https://api.github.com/repos/openclaw/openclaw/issues/67598 | Cron job payload.model override ignored — isolated sessions run on agent primary instead | ## Bug Report
**Version:** 2026.4.5
**Summary:** The `model` field set in a cron job's `payload` via `openclaw cron edit --model <model>` is not honored when the job runs. Isolated cron sessions always use the agent's primary model instead of the payload override.
## Steps to Reproduce
1. Create an isolated cron jo... | closed | completed | true | 1 | [] | [] | 2026-04-16T09:12:05Z | 2026-04-30T04:51:59Z | 2026-04-25T05:15:21Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | zachware | 57,385 | MDQ6VXNlcjU3Mzg1 | User | false |
openclaw/openclaw | 4,274,573,328 | I_kwDOQb6kR87-yNAQ | 67,600 | https://github.com/openclaw/openclaw/issues/67600 | https://api.github.com/repos/openclaw/openclaw/issues/67600 | [Bug]: tools.exec.timeoutSec is ignored when exec auto-backgrounds via yieldMs/background without explicit timeout | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
tools.exec.timeoutSec is documented as the default timeout for the exec tool, but it is not applied when an exec call uses yieldMs or background without an explicit timeout parameter. Foreground exec correctly ... | closed | completed | true | 2 | [
"bug",
"bug:behavior"
] | [] | 2026-04-16T09:18:33Z | 2026-04-30T04:35:08Z | 2026-04-27T10:06:40Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | dlmpx | 73,294,677 | MDQ6VXNlcjczMjk0Njc3 | User | false |
openclaw/openclaw | 4,274,642,418 | I_kwDOQb6kR87-yd3y | 67,602 | https://github.com/openclaw/openclaw/issues/67602 | https://api.github.com/repos/openclaw/openclaw/issues/67602 | [Bug]: Cron announce delivery fails with 'Unsupported channel: <channel-id>' for Slack channels | ## Description
Cron jobs with `--announce` delivery to Slack channels consistently fail with:
```
cron: failed to resolve failure destination target { error: 'Unsupported channel: c0at2q238mq' }
```
The cron job runs successfully (the agent executes and completes its task), but the announce delivery step fails to ro... | closed | completed | true | 3 | [] | [] | 2026-04-16T09:29:08Z | 2026-04-30T04:39:47Z | 2026-04-26T05:15:30Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | novaroagency | 215,420,112 | U_kgDODNcM0A | User | false |
openclaw/openclaw | 4,274,737,462 | I_kwDOQb6kR87-y1E2 | 67,606 | https://github.com/openclaw/openclaw/issues/67606 | https://api.github.com/repos/openclaw/openclaw/issues/67606 | Bug: Sending photos via Telegram crashes session with ollama/glm-5.1 (text-only primary model) — provider rejects format error | ## Bug Description
When sending a photo via Telegram to an agent whose primary model is text-only (e.g., `ollama/glm-5.1:cloud`), the entire session crashes with a provider format error. The user receives no response and the session becomes unresponsive until reset.
This is a specific, user-facing manifestation of tw... | closed | completed | true | 1 | [] | [] | 2026-04-16T09:45:43Z | 2026-04-28T04:52:47Z | 2026-04-25T18:13:16Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Joel-Claw | 274,497,724 | U_kgDOEFyAvA | User | false |
openclaw/openclaw | 4,274,741,682 | I_kwDOQb6kR87-y2Gy | 67,607 | https://github.com/openclaw/openclaw/issues/67607 | https://api.github.com/repos/openclaw/openclaw/issues/67607 | [Feature]: Include threadId in message envelope for thread replies on Slack | ### Summary
Include threadId in Slack thread message envelopes.
### Problem to solve
Agents receiving Slack thread messages have no way to identify thread context, the message envelope omits threadId, so agents can't distinguish thread replies from channel messages or fetch prior thread history. Users must compensat... | open | null | false | 1 | [
"enhancement"
] | [] | 2026-04-16T09:46:29Z | 2026-04-16T12:32:04Z | null | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | longinesequal | 239,612,679 | U_kgDODkgzBw | User | false |
openclaw/openclaw | 4,274,782,654 | I_kwDOQb6kR87-zAG- | 67,609 | https://github.com/openclaw/openclaw/issues/67609 | https://api.github.com/repos/openclaw/openclaw/issues/67609 | [Bug]: google-gemini-cli provider routes OAuth requests to generativelanguage.googleapis.com (403) instead of cloudcode-pa.googleapis.com | ### Bug type
Bug (incorrect behavior)
### Summary
The built-in `google-gemini-cli` provider sends OAuth-authenticated requests to `generativelanguage.googleapis.com/v1beta`, which rejects OAuth tokens with `403 ACCESS_TOKEN_SCOPE_INSUFFICIENT`. The correct endpoint for OAuth (Code Assist) tokens is `cloudcode-pa.goog... | closed | completed | true | 1 | [] | [] | 2026-04-16T09:53:43Z | 2026-04-28T04:52:46Z | 2026-04-25T18:13:31Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | bfloesser | 214,952,071 | U_kgDODM_ohw | User | false |
openclaw/openclaw | 4,274,787,371 | I_kwDOQb6kR87-zBQr | 67,610 | https://github.com/openclaw/openclaw/issues/67610 | https://api.github.com/repos/openclaw/openclaw/issues/67610 | [Bug]: requireSandboxBackendFactory throw error after sandbox plugin registered in openclaw starting stage | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
A new sandbox plugin is registered in openclaw starting stage, but requireSandboxBackendFactory throw error while send message to openclaw.
### Steps to reproduce
1. A new sandbox backend plugin, just named a... | open | null | false | 1 | [
"bug",
"bug:behavior"
] | [] | 2026-04-16T09:54:34Z | 2026-05-02T17:55:09Z | null | NONE | null | 20260502T180029Z | 2026-05-02T18:00:29Z | earon-han | 250,782,655 | U_kgDODvKjvw | User | false |
openclaw/openclaw | 4,274,791,191 | I_kwDOQb6kR87-zCMX | 67,611 | https://github.com/openclaw/openclaw/issues/67611 | https://api.github.com/repos/openclaw/openclaw/issues/67611 | [Feature]: Automatic Cleanup of Dreaming sessions. | ### Summary
Since I have enabled dreaming I can see a lot of dreaming sessions. I am not sure these should be seen as sessions clouding the session_list. Maybe remove them when complete? or create something different where they are are saved.
### Problem to solve
Agent_list containst the dreaming sessions. I thi... | closed | completed | true | 1 | [
"enhancement"
] | [] | 2026-04-16T09:55:16Z | 2026-04-28T04:52:45Z | 2026-04-25T18:13:38Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | najef1979-code | 233,555,716 | U_kgDODevHBA | User | false |
openclaw/openclaw | 4,274,894,187 | I_kwDOQb6kR87-zbVr | 67,615 | https://github.com/openclaw/openclaw/issues/67615 | https://api.github.com/repos/openclaw/openclaw/issues/67615 | Heartbeat mechanism interrupts ongoing task processes | ## Description
The heartbeat/heartbeat poll mechanism sometimes interrupts or conflicts with ongoing task processes, causing tasks to be prematurely terminated or interrupted.
## Reproduction
1. Start a long-running task (e.g., RSS news scraping + analysis pipeline that takes several minutes)
2. A heartbeat poll arr... | closed | completed | true | 1 | [] | [] | 2026-04-16T10:10:13Z | 2026-04-28T04:52:45Z | 2026-04-25T18:13:45Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | jonah791 | 175,467,022 | U_kgDOCnVqDg | User | false |
openclaw/openclaw | 4,274,963,927 | I_kwDOQb6kR87-zsXX | 67,617 | https://github.com/openclaw/openclaw/issues/67617 | https://api.github.com/repos/openclaw/openclaw/issues/67617 | Expose telegram_file_id in media context for stickers | # Feature Request: Expose `telegram_file_id` in media context for stickers
## Use Case
Building a sticker lexicon skill that stores and retrieves Telegram stickers. Currently stickers arrive as downloaded files, but the original `file_id` needed to resend them as native stickers is lost.
## Current Behavior
- Sticker... | closed | completed | true | 1 | [] | [] | 2026-04-16T10:21:11Z | 2026-04-28T04:52:44Z | 2026-04-25T18:13:52Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Souped-code | 225,992,359 | U_kgDODXhepw | User | false |
openclaw/openclaw | 4,274,971,722 | I_kwDOQb6kR87-zuRK | 67,618 | https://github.com/openclaw/openclaw/issues/67618 | https://api.github.com/repos/openclaw/openclaw/issues/67618 | Improve pairing required error message to include approval command | ## Problem
When a device needs pairing approval after an upgrade, the error message only shows:
```
Pairing required. Run `openclaw devices list`, approve your request ID, then reconnect.
```
This tells users to list devices, but **doesn't explain how to approve**. Users are left confused about what command to run a... | open | null | false | 2 | [] | [] | 2026-04-16T10:22:29Z | 2026-05-02T23:12:34Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | xyzliao | 223,995,557 | U_kgDODVnmpQ | User | false |
openclaw/openclaw | 4,274,987,648 | I_kwDOQb6kR87-zyKA | 67,619 | https://github.com/openclaw/openclaw/issues/67619 | https://api.github.com/repos/openclaw/openclaw/issues/67619 | [Bug]: Agent response context lost after switching between chat sessions without explicit save | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
When switching between active chat sessions without explicitly saving, the previous session's AI agent context and conversation history is lost and cannot be recovered.
### Steps to reproduce
1. Open OpenClaw with 2 or mor... | closed | not_planned | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-16T10:25:20Z | 2026-04-28T04:52:43Z | 2026-04-25T18:13:59Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Vansh1811 | 192,606,842 | U_kgDOC3ryeg | User | false |
openclaw/openclaw | 4,275,011,327 | I_kwDOQb6kR87-z37_ | 67,621 | https://github.com/openclaw/openclaw/issues/67621 | https://api.github.com/repos/openclaw/openclaw/issues/67621 | [Bug] Exec preflight flags $ in Python string literals as shell variable injection | ## Summary
The exec preflight source-file scanner (`security=full`) flags `$IDENTIFIER` patterns inside Python string literals (docstrings, f-strings, argparse help text) as shell variable injection. This blocks execution of any Python skill that prints dollar-formatted output or uses `$` as a currency label.
A simpl... | open | null | false | 2 | [] | [] | 2026-04-16T10:29:16Z | 2026-05-03T07:44:50Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | adlai88 | 6,263,940 | MDQ6VXNlcjYyNjM5NDA= | User | false |
openclaw/openclaw | 4,275,016,971 | I_kwDOQb6kR87-z5UL | 67,622 | https://github.com/openclaw/openclaw/issues/67622 | https://api.github.com/repos/openclaw/openclaw/issues/67622 | [Feature Request] Support file/image upload in browser Control UI (webchat) | ## Use Case
I need to use OpenClaw browser Control UI (http://127.0.0.1:18789/) as my primary work interface because:
1. Large screen - much better for viewing documents, screenshots, and code
2. Multi-window - can reference materials while chatting
3. Already my daily workflow
## Problem
Currently the browser Contr... | closed | not_planned | true | 3 | [
"clownfish"
] | [
"vincentkoc"
] | 2026-04-16T10:30:11Z | 2026-04-28T04:32:11Z | 2026-04-26T03:03:57Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | biteerqiongen-hash | 275,756,476 | U_kgDOEG-1vA | User | false |
openclaw/openclaw | 4,275,028,106 | I_kwDOQb6kR87-z8CK | 67,624 | https://github.com/openclaw/openclaw/issues/67624 | https://api.github.com/repos/openclaw/openclaw/issues/67624 | [Bug]: openai-codex/gpt-5.4 returns HTML payload instead of model output (webchat/TUI hangs, messages queue) | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Issue with webchat where Codex route returns HTML in `outputs[0].text` causing webchat and TUI requests to hang and queue.
### Steps to reproduce
run `openclaw infer model run --local --model "openai-codex/gpt-5.4" --promp... | closed | completed | true | 3 | [
"bug",
"regression"
] | [] | 2026-04-16T10:31:53Z | 2026-04-24T04:26:29Z | 2026-04-21T06:40:09Z | NONE | null | 20260424T060021Z | 2026-04-24T06:00:21Z | theglove44 | 148,962,746 | U_kgDOCOD9ug | User | false |
openclaw/openclaw | 4,275,128,612 | I_kwDOQb6kR87-0Ukk | 67,625 | https://github.com/openclaw/openclaw/issues/67625 | https://api.github.com/repos/openclaw/openclaw/issues/67625 | [Bug]: Billing error message fires after billing issue is resolved — deduplication missing | ## Summary
The billing error notification fires again after the underlying billing issue has already been resolved, causing user confusion. Related to (now-locked) #15055 but adds a new finding: **deduplication**.
## Environment
- OpenClaw: 2026.4.14
- OS: Ubuntu 24.04.4 LTS (ARM64)
- Provider: Anthropic, claude-son... | closed | completed | true | 2 | [] | [] | 2026-04-16T10:48:25Z | 2026-04-28T04:33:04Z | 2026-04-25T18:14:06Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | doronski | 24,498,888 | MDQ6VXNlcjI0NDk4ODg4 | User | false |
openclaw/openclaw | 4,275,130,275 | I_kwDOQb6kR87-0U-j | 67,626 | https://github.com/openclaw/openclaw/issues/67626 | https://api.github.com/repos/openclaw/openclaw/issues/67626 | [Bug]: 越更新越更的像屎一样 | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
以前引导式配置完，几乎直接可以用tui，现在各种配置，用codex和gemini改了半天，都用不了。三天两头的改引导配置和json配置，装了半天，一直有问题，反倒是clawbot时期第一次安装就成功。现在是一坨大便！
### Steps to reproduce
以前引导式配置完，几乎直接可以用tui，现在各种配置，用codex和gemini改了半天，都用不了。三天两头的改引导配置和json配置，装了半天，一直有问题，反倒是clawbot时... | open | null | false | 3 | [
"bug",
"regression"
] | [] | 2026-04-16T10:48:43Z | 2026-05-03T07:44:51Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | 285878535 | 25,810,268 | MDQ6VXNlcjI1ODEwMjY4 | User | false |
openclaw/openclaw | 4,275,140,059 | I_kwDOQb6kR87-0XXb | 67,629 | https://github.com/openclaw/openclaw/issues/67629 | https://api.github.com/repos/openclaw/openclaw/issues/67629 | [Feature]: I want each timed task cron to show if it is running or not, add an identifier | ### Summary
I want each timed task cron to show if it is running or not, add an identifier
### Problem to solve
I want each timed task cron to show if it is running or not, add an identifier
### Proposed solution
I want each timed task cron to show if it is running or not, add an identifier
### Alternatives consi... | open | null | false | 1 | [
"enhancement"
] | [] | 2026-04-16T10:50:28Z | 2026-05-03T07:44:52Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | cyrus123456 | 73,327,722 | MDQ6VXNlcjczMzI3NzIy | User | false |
openclaw/openclaw | 4,275,145,807 | I_kwDOQb6kR87-0YxP | 67,631 | https://github.com/openclaw/openclaw/issues/67631 | https://api.github.com/repos/openclaw/openclaw/issues/67631 | [Bug]: Mac App and npm CLI both install same LaunchAgent label causing duplicate gateway processes | ## Bug type
Bug
## Summary
On macOS, installing the OpenClaw Mac App and running `openclaw doctor --fix` (which reinstalls the npm LaunchAgent) causes two concurrent gateway processes to run with the same LaunchAgent label `ai.openclaw.gateway`, leading to port conflicts and broken RPC between Mac App and Telegram.
#... | open | null | false | 3 | [] | [] | 2026-04-16T10:51:26Z | 2026-05-03T07:44:53Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | njbilly007-lab | 259,042,561 | U_kgDOD3CtAQ | User | false |
openclaw/openclaw | 4,275,236,359 | I_kwDOQb6kR87-0u4H | 67,636 | https://github.com/openclaw/openclaw/issues/67636 | https://api.github.com/repos/openclaw/openclaw/issues/67636 | feat: cost-observer skill — per-agent/channel/user spend tracking | ## Summary
Track model spend per agent, channel, and user across all OpenClaw sessions. Competitive gap vs Hive and VoltAgent which both ship cost enforcement natively.
## Motivation
Hive (10.2k GitHub stars) ships with budget enforcement per agent graph. VoltAgent (7.2k stars) includes cost tracking as a first-clas... | closed | not_planned | true | 1 | [] | [] | 2026-04-16T11:08:01Z | 2026-04-28T04:52:43Z | 2026-04-25T18:14:13Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | finneyai-app[bot] | 245,033,325 | BOT_kgDODprpbQ | Bot | false |
openclaw/openclaw | 4,275,248,060 | I_kwDOQb6kR87-0xu8 | 67,637 | https://github.com/openclaw/openclaw/issues/67637 | https://api.github.com/repos/openclaw/openclaw/issues/67637 | [qa-lab] Add tool-call trace inspection to QaBusMessage for richer scenario assertions | ## Problem
QaBusMessage (extensions/qa-channel/src/protocol.ts) only carries text payloads. Scenarios cannot assert on which tools were called, only on response text content. This makes behavioral parity scenarios (e.g., verifying GPT-5.4 actually calls `exec` instead of answering from memory) unreliable — the asserti... | open | null | false | 2 | [] | [] | 2026-04-16T11:09:50Z | 2026-05-02T23:57:43Z | null | CONTRIBUTOR | null | 20260503T000641Z | 2026-05-03T00:06:41Z | 100yenadmin | 239,388,517 | U_kgDODkTHZQ | User | false |
openclaw/openclaw | 4,275,289,034 | I_kwDOQb6kR87-07vK | 67,638 | https://github.com/openclaw/openclaw/issues/67638 | https://api.github.com/repos/openclaw/openclaw/issues/67638 | [Bug]: Idle persistent ACP sessions still consume maxConcurrentSessions and block later spawn | ## Summary
Persistent ACP sessions appear to keep occupying `acp.maxConcurrentSessions` after a run has finished and the session has returned to `idle`. In that state, a later `sessions_spawn` for the same agent/workflow fails with:
```text
ACP max concurrent sessions reached (1/1)
```
This creates a dead-end for au... | closed | completed | true | 1 | [] | [] | 2026-04-16T11:16:17Z | 2026-04-19T04:23:17Z | 2026-04-16T11:18:49Z | NONE | null | 20260419T052035Z | 2026-04-19T05:20:35Z | XanderZhu | 25,664,794 | MDQ6VXNlcjI1NjY0Nzk0 | User | false |
openclaw/openclaw | 4,275,378,991 | I_kwDOQb6kR87-1Rsv | 67,640 | https://github.com/openclaw/openclaw/issues/67640 | https://api.github.com/repos/openclaw/openclaw/issues/67640 | [Feature]: Add bundled YuanBao (元宝) channel plugin for Tencent YuanBao Bot integration | ### Summary
OpenClaw currently has no built-in channel plugin for Tencent YuanBao (元宝) — a widely used AI assistant and IM bot platform in China. Users who want to connect their OpenClaw agents to YuanBao bots must build custom integrations from scratch, dealing with YuanBao's proprietary Protobuf-over-WebSocket proto... | closed | not_planned | true | 4 | [
"enhancement"
] | [
"sliverp"
] | 2026-04-16T11:30:03Z | 2026-04-30T04:39:47Z | 2026-04-26T05:15:35Z | CONTRIBUTOR | null | 20260430T060018Z | 2026-04-30T06:00:18Z | loongfay | 16,909,883 | MDQ6VXNlcjE2OTA5ODgz | User | false |
openclaw/openclaw | 4,275,428,716 | I_kwDOQb6kR87-1d1s | 67,646 | https://github.com/openclaw/openclaw/issues/67646 | https://api.github.com/repos/openclaw/openclaw/issues/67646 | [Bug]: OpenRouter text-inference models return silent empty responses through cron-agent-turn harness in OpenClaw 2026.4.14 | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Environment:
- OpenClaw 2026.4.14 (323493f)
- macOS / Homebrew install
- OpenRouter provider plugin loaded
- OpenRouter auth valid and paid credit present
Problem:
When OpenRouter text-inference models are used through the ... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-16T11:37:22Z | 2026-04-21T04:22:00Z | 2026-04-18T16:02:15Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | RocksNStocks | 84,497,323 | MDQ6VXNlcjg0NDk3MzIz | User | false |
openclaw/openclaw | 4,275,438,600 | I_kwDOQb6kR87-1gQI | 67,648 | https://github.com/openclaw/openclaw/issues/67648 | https://api.github.com/repos/openclaw/openclaw/issues/67648 | [Bug]: SKILL.md changes don't auto-execute - category-mapping.json not read | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
## Problem Description
When modifying SKILL.md or creating configuration files like `category-mapping.json`, the changes don't actually affect the agent's behavior.
## Root Cause
SKILL.md files are documentation/prompts, ... | closed | not_planned | true | 2 | [
"bug",
"regression"
] | [] | 2026-04-16T11:39:06Z | 2026-04-28T04:52:42Z | 2026-04-25T18:14:28Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | wzhgfwy001 | 275,329,920 | U_kgDOEGkzgA | User | false |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.