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 | 4,347,163,456 | I_kwDOQb6kR88AAAABAxxzQA | 73,898 | https://github.com/openclaw/openclaw/issues/73898 | https://api.github.com/repos/openclaw/openclaw/issues/73898 | [Feature]: Control app option to auto-play assistant TTS audio | ### Summary
Add a Control app / Control UI setting to automatically play assistant TTS audio attachments when they are generated.
### Problem to solve
When `messages.tts.auto` is set to `always`, OpenClaw correctly generates MP3 audio for assistant replies, but in the Control app the audio appears as an embedded pl... | open | null | false | 2 | [
"enhancement"
] | [] | 2026-04-29T00:48:34Z | 2026-05-03T07:45:32Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | zachisfine | 131,436,334 | U_kgDOB9WPLg | User | false |
openclaw/openclaw | 4,347,240,832 | I_kwDOQb6kR88AAAABAx2hgA | 73,909 | https://github.com/openclaw/openclaw/issues/73909 | https://api.github.com/repos/openclaw/openclaw/issues/73909 | [Bug]: Ollama cloud models (gemini-3-flash-preview:cloud, deepseek-v4-pro:cloud) fail with "Assistant turn failed before producing content" or silent fallback | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Two Ollama cloud models that work perfectly via direct Ollama API call fail when used through OpenClaw:
1. ollama/gemini-3-flash-preview:cloud → "Assistant turn failed before producing content"
2. ollama/depse... | closed | completed | true | 3 | [
"bug",
"bug:behavior"
] | [] | 2026-04-29T01:14:07Z | 2026-05-01T04:39:09Z | 2026-04-29T04:14:58Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | chtse53 | 6,392,925 | MDQ6VXNlcjYzOTI5MjU= | User | false |
openclaw/openclaw | 4,347,277,459 | I_kwDOQb6kR88AAAABAx4wkw | 73,916 | https://github.com/openclaw/openclaw/issues/73916 | https://api.github.com/repos/openclaw/openclaw/issues/73916 | [Bug]: Gateway binds port but hangs / never responds on Intel macOS with existing ~/.openclaw state | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
OpenClaw gateway starts, binds `127.0.0.1:18789`, and logs `gateway ready`, but HTTP/WS never respond. `curl` connects then times out with 0 bytes received. CPU stays high.
A clean temporary config/state on port `18790` wor... | open | null | false | 3 | [
"bug"
] | [] | 2026-04-29T01:27:59Z | 2026-05-03T06:50:44Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | AKKarthikeyan | 94,901,080 | U_kgDOBagTWA | User | false |
openclaw/openclaw | 4,347,353,311 | I_kwDOQb6kR88AAAABAx9Y3w | 73,939 | https://github.com/openclaw/openclaw/issues/73939 | https://api.github.com/repos/openclaw/openclaw/issues/73939 | [v1.4.9] skills.entries custom fields rejected by schema validation at runtime | ## Bug Description
`skills.entries.<skill-name>` 下自定义字段（如 `apiKey`、`api_endpoint`）在 `openclaw.json` 写入时通过了 schema 校验，但在 Gateway reload 时被 runtime validation 拒绝，导致配置回滚到备份版本。
**受影响场景**：
- `skills.entries.tavily` 的 `apiKey` 字段
- `skills.entries.kaiwu8` 的 `api_endpoint` + `user_api_key` 字段
**Schema 现状**：
- `skills.entri... | closed | not_planned | true | 2 | [] | [] | 2026-04-29T01:55:56Z | 2026-05-03T04:34:30Z | 2026-05-01T01:14:09Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | MarkMa84 | 19,796,416 | MDQ6VXNlcjE5Nzk2NDE2 | User | false |
openclaw/openclaw | 4,347,355,604 | I_kwDOQb6kR88AAAABAx9h1A | 73,941 | https://github.com/openclaw/openclaw/issues/73941 | https://api.github.com/repos/openclaw/openclaw/issues/73941 | [Feature Request] Auto-remove spaces between Chinese and English characters in file paths | ## Problem Description
When AI models (such as Qwen, PAI, MiniMax, etc.) output file paths, they automatically add spaces between Chinese and English characters.
### Example
**Expected path** (actual directory name):
`
Z:\G营销\AI临时文件\report.xlsx
`
**Model output** (incorrect):
`
Z:\G 营销\AI 临时文件\report.xlsx
`
### Im... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T01:56:34Z | 2026-05-03T04:36:29Z | 2026-04-30T14:59:57Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | visionzhang-prog | 266,700,237 | U_kgDOD-WFzQ | User | false |
openclaw/openclaw | 4,347,026,020 | I_kwDOQb6kR88AAAABAxpaZA | 73,875 | https://github.com/openclaw/openclaw/issues/73875 | https://api.github.com/repos/openclaw/openclaw/issues/73875 | BUG: ACP runtime sends unsupported config options to Claude ACP adapter | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
OpenClaw-managed ACP sessions for Claude fail before producing output when OpenClaw applies generic runtime options as ACP session config options. Direct ACPX to Claude works on the same system, while the OpenC... | open | null | false | 1 | [
"bug"
] | [] | 2026-04-29T00:04:50Z | 2026-05-03T06:07:50Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | kAIborg24 | 261,855,322 | U_kgDOD5uYWg | User | false |
openclaw/openclaw | 4,347,028,609 | I_kwDOQb6kR88AAAABAxpkgQ | 73,876 | https://github.com/openclaw/openclaw/issues/73876 | https://api.github.com/repos/openclaw/openclaw/issues/73876 | [Bug]: Direct provider names inconsistency (Eg. 'moonshot' should be 'moonshotai' to match OpenRouter slug convention) | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
The direct Moonshot AI provider is named moonshot in openclaw, but OpenRouter (and Moonshot AI's own branding) uses moonshotai as the organization slug. This creates a confusing and inconsistent naming mismatch for users who... | open | null | false | 2 | [
"bug",
"regression"
] | [] | 2026-04-29T00:05:44Z | 2026-05-04T18:47:47Z | null | NONE | null | 20260504T210031Z | 2026-05-04T21:00:31Z | jimdawdy-hub | 262,052,366 | U_kgDOD56aDg | User | false |
openclaw/openclaw | 4,347,036,810 | I_kwDOQb6kR88AAAABAxqEig | 73,878 | https://github.com/openclaw/openclaw/issues/73878 | https://api.github.com/repos/openclaw/openclaw/issues/73878 | infer model run --gateway fails 'No callable tools remain' when modelRun:true + tools.allow has explicit entries (4.25) | # Upstream bug report DRAFT — pending Trent blind-pass
**Target:** github.com/openclaw/openclaw/issues
**Status:** DRAFT v0 — not filed. Needs Trent review of repro + RCA before submission.
---
## Title
`infer model run --gateway` fails with "No callable tools remain" when `modelRun:true` + `tools.allow` has explic... | closed | not_planned | true | 3 | [] | [] | 2026-04-29T00:08:25Z | 2026-05-02T04:29:11Z | 2026-04-29T05:48:12Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | islandpreneur007 | 258,422,617 | U_kgDOD2c3WQ | User | false |
openclaw/openclaw | 4,347,365,511 | I_kwDOQb6kR88AAAABAx-Ihw | 73,951 | https://github.com/openclaw/openclaw/issues/73951 | https://api.github.com/repos/openclaw/openclaw/issues/73951 | Bug Report: replyInThread config 'disabled' not working | # Bug Report: replyInThread config 'disabled' not working
**Repo:** openclaw/openclaw
**Title:** replyInThread config 'disabled' not working - thread replies still sent
---
**Description**
When setting `replyInThread: "disabled"` in the Feishu channel account config, thread replies are still being sent. This causes... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T01:59:48Z | 2026-05-03T04:35:05Z | 2026-04-30T20:37:35Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | GQ-GH | 242,907,409 | U_kgDODnp5EQ | User | false |
openclaw/openclaw | 4,347,391,100 | I_kwDOQb6kR88AAAABAx_sfA | 73,956 | https://github.com/openclaw/openclaw/issues/73956 | https://api.github.com/repos/openclaw/openclaw/issues/73956 | [Bug]: Control UI duplicates assistant replies via assistant-media delivery path | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Control UI/WebChat is showing duplicate assistant replies. Backend session history shows a normal final assistant message followed by a second gateway-injected assistant message with idempotencyKey ending in `:... | closed | completed | true | 1 | [
"bug"
] | [] | 2026-04-29T02:07:58Z | 2026-05-05T04:28:43Z | 2026-05-02T05:22:04Z | NONE | null | 20260505T060034Z | 2026-05-05T06:00:34Z | zachisfine | 131,436,334 | U_kgDOB9WPLg | User | false |
openclaw/openclaw | 4,347,270,419 | I_kwDOQb6kR88AAAABAx4VEw | 73,914 | https://github.com/openclaw/openclaw/issues/73914 | https://api.github.com/repos/openclaw/openclaw/issues/73914 | [Bug]: WhatsApp inbound handler dead after 408 reconnect cycle — watchdog detects but reconnect doesn't restore (2026.4.26 regression) | ## Summary
After upgrading to 2026.4.26 (which included WhatsApp stability improvements), the WhatsApp inbound handler enters a silent failure loop within <24 hours. The built-in watchdog correctly detects the problem (`app-silent`) but the 499 reconnect it triggers does not restore inbound message delivery. This pers... | closed | completed | true | 2 | [] | [] | 2026-04-29T01:25:14Z | 2026-05-03T04:35:59Z | 2026-04-30T16:18:36Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | legonhilltech-jpg | 272,396,537 | U_kgDOEDxw-Q | User | false |
openclaw/openclaw | 4,347,407,167 | I_kwDOQb6kR88AAAABAyArPw | 73,957 | https://github.com/openclaw/openclaw/issues/73957 | https://api.github.com/repos/openclaw/openclaw/issues/73957 | [Feature Request] Expose MiniMax ASR (Speech-to-Text) API in minimax plugin | ## Feature Description
MiniMax provides ASR (Automatic Speech Recognition) API capabilities, but the current OpenClaw minimax plugin only exposes text chat models (M2.7, M2.7-highspeed) and does not expose the ASR endpoint.
## Expected Behavior
The minimax plugin should expose MiniMax's speech-to-text ASR capabiliti... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T02:12:26Z | 2026-05-03T04:34:30Z | 2026-05-01T01:14:15Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | ulnacranium17 | 257,941,027 | U_kgDOD1_eIw | User | false |
openclaw/openclaw | 4,347,422,979 | I_kwDOQb6kR88AAAABAyBpAw | 73,963 | https://github.com/openclaw/openclaw/issues/73963 | https://api.github.com/repos/openclaw/openclaw/issues/73963 | [Bug]: 2026.4.26 injects unsupported metadata field into openai-codex Responses requests; ChatGPT backend returns HTTP 400, surfaced as "400 status code (no body)" | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
After upgrading to 2026.4.26, every request the `openai-codex` provider
sends to `https://chatgpt.com/backend-api/codex/responses` is rejected by
the ChatGPT codex backend with HTTP 400
`{"detail":"Unsupported parameter: met... | closed | completed | true | 2 | [] | [] | 2026-04-29T02:17:51Z | 2026-05-03T04:35:22Z | 2026-04-30T19:18:49Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | kasanuowa | 23,185,829 | MDQ6VXNlcjIzMTg1ODI5 | User | false |
openclaw/openclaw | 4,347,427,170 | I_kwDOQb6kR88AAAABAyB5Yg | 73,966 | https://github.com/openclaw/openclaw/issues/73966 | https://api.github.com/repos/openclaw/openclaw/issues/73966 | [Bug]: Slack message-send action drops resolved config, causing SecretRef-backed botToken to appear missing | ## Bug type
Regression / behavior bug
## Summary
`openclaw message send --channel slack ...` fails with:
`Error: Slack bot token missing for account "default" (set channels.slack.accounts.default.botToken or SLACK_BOT_TOKEN for default).`
This happens even when Slack is otherwise configured correctly and `openclaw... | closed | completed | true | 2 | [] | [] | 2026-04-29T02:19:31Z | 2026-05-03T04:35:50Z | 2026-04-30T17:05:01Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | slatem | 1,423,703 | MDQ6VXNlcjE0MjM3MDM= | User | false |
openclaw/openclaw | 4,347,036,954 | I_kwDOQb6kR88AAAABAxqFGg | 73,880 | https://github.com/openclaw/openclaw/issues/73880 | https://api.github.com/repos/openclaw/openclaw/issues/73880 | openclaw update exits code 1 on disabled-plugin ClawHub 429 even though core update succeeded; should treat disabled plugins as no-op for sync | # Upstream bug report DRAFT — pending Trent blind-pass
**Target:** github.com/openclaw/openclaw/issues
**Status:** DRAFT v0 — not filed.
---
## Title
`openclaw update` exits code 1 on disabled-plugin ClawHub 429 even though core update succeeded; should treat disabled plugins as no-op for sync
## Environment
- Op... | closed | completed | true | 2 | [] | [] | 2026-04-29T00:08:29Z | 2026-05-01T04:40:51Z | 2026-04-29T02:36:13Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | islandpreneur007 | 258,422,617 | U_kgDOD2c3WQ | User | false |
openclaw/openclaw | 4,347,444,273 | I_kwDOQb6kR88AAAABAyC8MQ | 73,977 | https://github.com/openclaw/openclaw/issues/73977 | https://api.github.com/repos/openclaw/openclaw/issues/73977 | ACP Codex auth regression: acpx wrapper forces isolated CODEX_HOME and breaks reuse of valid Codex OAuth | # ACP Codex auth regression: acpx wrapper forces isolated `CODEX_HOME` and breaks reuse of valid Codex OAuth
## Summary
`sessions_spawn(runtime="acp", agentId="codex")` can fail authentication even when Codex CLI is already logged in and working on the same machine, because the OpenClaw ACPX wrapper forces `CODEX_HOM... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T02:25:58Z | 2026-05-01T04:40:50Z | 2026-04-29T02:37:19Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | f1yankees | 32,306,781 | MDQ6VXNlcjMyMzA2Nzgx | User | false |
openclaw/openclaw | 4,347,432,822 | I_kwDOQb6kR88AAAABAyCPdg | 73,971 | https://github.com/openclaw/openclaw/issues/73971 | https://api.github.com/repos/openclaw/openclaw/issues/73971 | [Feature]: 在页面上新增其他可用大模型的序号和实际报错不一致 | ### Summary
版本:Version v2026.4.26
<img width="2212" height="1325" alt="Image" src="https://github.com/user-attachments/assets/823a9c82-00e3-4c27-8846-fd521a32f568" />
我的是#4前端的配置有问题,但是报错显示的是第三个.该错误不影响功能,但是影响体验,代码从0计数,但是前端显示的是#4,容易让人误解
### Problem to solve
建议前端和后端显示一样,建议写为一样的.
### Proposed solution
建议前端和后端显示一样,建议写... | open | null | false | 1 | [
"enhancement"
] | [] | 2026-04-29T02:21:44Z | 2026-05-03T11:50:10Z | null | NONE | null | 20260503T120038Z | 2026-05-03T12:00:38Z | qinqing133 | 76,898,703 | MDQ6VXNlcjc2ODk4NzAz | User | false |
openclaw/openclaw | 4,347,469,827 | I_kwDOQb6kR88AAAABAyEgAw | 73,990 | https://github.com/openclaw/openclaw/issues/73990 | https://api.github.com/repos/openclaw/openclaw/issues/73990 | Bug: local Llama/OpenAI-compatible sessions can show 0% context when usage telemetry is missing | ## Summary
Local Llama / OpenAI-compatible local models can still show incorrect context usage (`0%`, missing usage, or stale low usage) when the model response does not provide provider usage telemetry. The status/session usage path can return no usable snapshot instead of estimating usage from the transcript.
This ... | open | null | false | 2 | [] | [] | 2026-04-29T02:35:01Z | 2026-05-03T12:31:55Z | null | NONE | null | 20260503T150031Z | 2026-05-03T15:00:31Z | ulttla | 19,655,702 | MDQ6VXNlcjE5NjU1NzAy | User | false |
openclaw/openclaw | 4,347,445,064 | I_kwDOQb6kR88AAAABAyC_SA | 73,978 | https://github.com/openclaw/openclaw/issues/73978 | https://api.github.com/repos/openclaw/openclaw/issues/73978 | [Bug] Discord plugin slash commands miss 3s ack deadline (regression of #60880 for plugin-registered commands) — "Unknown interaction" | ### Description
On OpenClaw `2026.4.26`, slash commands provided by plugins (`/vc join`, `/voice`, `/codex`, `/claude_mem_*`, etc.) consistently fail with **"Unknown interaction"** or **"Interaction has already been acknowledged"** when invoked from the Discord client.
**Likely a regression** of #55240 / #58602 / #60... | open | null | false | 2 | [] | [] | 2026-04-29T02:26:17Z | 2026-05-05T09:33:07Z | null | NONE | null | 20260505T120033Z | 2026-05-05T12:00:33Z | kzicherman | 12,486,923 | MDQ6VXNlcjEyNDg2OTIz | User | false |
openclaw/openclaw | 4,347,379,791 | I_kwDOQb6kR88AAAABAx_ATw | 73,954 | https://github.com/openclaw/openclaw/issues/73954 | https://api.github.com/repos/openclaw/openclaw/issues/73954 | [Bug]: Agent outputs /approve <shell-command> instead of /approve <slug> <decision> | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
When exec approval is required, the agent outputs `/approve <full-shell-command>` (the literal command it wants to run) instead of the correct `/approve <slug> allow-once|allow-always|deny` format from the tool... | closed | completed | true | 3 | [
"bug"
] | [] | 2026-04-29T02:04:26Z | 2026-05-02T04:29:19Z | 2026-04-29T05:45:28Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | guarismo | 33,353,023 | MDQ6VXNlcjMzMzUzMDIz | User | false |
openclaw/openclaw | 4,347,356,521 | I_kwDOQb6kR88AAAABAx9laQ | 73,944 | https://github.com/openclaw/openclaw/issues/73944 | https://api.github.com/repos/openclaw/openclaw/issues/73944 | [Feature Request] Auto-remove spaces between Chinese and English characters in file paths | ## Problem Description
When AI models (such as Qwen, PAI, etc.) output file paths, they automatically add spaces between Chinese and English characters.
### Example
**Expected path** (actual directory name):
`
Z:\G营销\AI临时文件\report.xlsx
`
**Model output** (incorrect):
`
Z:\G 营销\AI 临时文件\report.xlsx
`
### Impact
1. ... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T01:56:47Z | 2026-05-01T04:40:47Z | 2026-04-29T02:41:12Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | visionzhang-prog | 266,700,237 | U_kgDOD-WFzQ | User | false |
openclaw/openclaw | 4,347,478,995 | I_kwDOQb6kR88AAAABAyFD0w | 73,992 | https://github.com/openclaw/openclaw/issues/73992 | https://api.github.com/repos/openclaw/openclaw/issues/73992 | [Bug]: `openclaw nodes list` intermittently times out on local loopback and can wedge follow-up gateway probe (2026.4.25) | ### Bug type
Behavior/performance bug
### Summary
`openclaw nodes list` intermittently times out against a healthy-looking local loopback gateway on 2026.4.25, and in some failure windows the timeout also spills over into `openclaw gateway probe`, suggesting `nodes.list` can still wedge or starve gateway RPC handlin... | closed | completed | true | 1 | [] | [] | 2026-04-29T02:37:28Z | 2026-05-03T04:35:05Z | 2026-04-30T20:37:43Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | vrurg | 1,586,992 | MDQ6VXNlcjE1ODY5OTI= | User | false |
openclaw/openclaw | 4,347,486,167 | I_kwDOQb6kR88AAAABAyFf1w | 73,997 | https://github.com/openclaw/openclaw/issues/73997 | https://api.github.com/repos/openclaw/openclaw/issues/73997 | [Feature]: WhatsApp dmPolicy "observe" mode — ingest DMs as read-only log without triggering agent turns | ## Summary
Add a new `dmPolicy: "observe"` option for WhatsApp accounts that ingests incoming DMs into a structured read-only log but **never routes them to an agent session** and **never triggers an agent turn**.
## Use Case
In a multi-agent setup, agents need **contextual awareness** of what is happening on the pr... | closed | not_planned | true | 2 | [] | [] | 2026-04-29T02:39:07Z | 2026-05-04T04:36:16Z | 2026-05-01T14:20:21Z | NONE | null | 20260504T060031Z | 2026-05-04T06:00:31Z | legonhilltech-jpg | 272,396,537 | U_kgDOEDxw-Q | User | false |
openclaw/openclaw | 4,347,502,340 | I_kwDOQb6kR88AAAABAyGfBA | 74,003 | https://github.com/openclaw/openclaw/issues/74003 | https://api.github.com/repos/openclaw/openclaw/issues/74003 | [Bug]: before_tool_call plugin approval fails with 'no approval route' — turnSourceChannel not passed to plugin.approval.request | ## Bug Description
When a `before_tool_call` hook returns `requireApproval`, the plugin approval flow fails with:
```
Plugin approval unavailable (no approval route)
```
The approval cannot be routed to the originating channel (e.g. Telegram) because the routing metadata is never passed to `plugin.approval.request`.... | open | null | false | 2 | [] | [] | 2026-04-29T02:43:32Z | 2026-05-03T05:00:02Z | null | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | choury | 4,332,542 | MDQ6VXNlcjQzMzI1NDI= | User | false |
openclaw/openclaw | 4,347,516,747 | I_kwDOQb6kR88AAAABAyHXSw | 74,006 | https://github.com/openclaw/openclaw/issues/74006 | https://api.github.com/repos/openclaw/openclaw/issues/74006 | [Bug]: gateway always failed after openclaw version updated | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
I’m wondering if I’m the only one running into this issue. Why do I encounter so many problems every time I update OpenClaw?
Previous versions were mostly fine, and I could usually fix things with doctor --fix or --repair. ... | closed | completed | true | 3 | [
"bug",
"regression"
] | [] | 2026-04-29T02:47:11Z | 2026-05-03T04:37:53Z | 2026-04-30T08:59:21Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | louisice | 29,276,312 | MDQ6VXNlcjI5Mjc2MzEy | User | false |
openclaw/openclaw | 4,347,516,488 | I_kwDOQb6kR88AAAABAyHWSA | 74,005 | https://github.com/openclaw/openclaw/issues/74005 | https://api.github.com/repos/openclaw/openclaw/issues/74005 | [Bug]: openai-completions to api.openai.com bypasses env-proxy dispatcher in HTTPS_PROXY-only sandbox; xAI / Anthropic / Discord on the same gateway PID work fine (2026.4.26) |
### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
In a sandbox where the **only** route to the public internet is an HTTPS forward proxy (`HTTPS_PROXY=http://<proxy>:3128`), gateway requests to `api.openai.com` via `openai-completions` fail every turn with `L... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T02:47:08Z | 2026-05-04T04:34:17Z | 2026-05-02T03:18:57Z | NONE | null | 20260504T060031Z | 2026-05-04T06:00:31Z | HeilbronAILabs | 274,369,335 | U_kgDOEFqLNw | User | false |
openclaw/openclaw | 4,347,549,013 | I_kwDOQb6kR88AAAABAyJVVQ | 74,014 | https://github.com/openclaw/openclaw/issues/74014 | https://api.github.com/repos/openclaw/openclaw/issues/74014 | [Bug]: Telegram channel in v2026.4.26 ignores HTTP_PROXY env vars | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Title: [Bug] Telegram channel in v2026.4.26 ignores HTTP_PROXY/HTTPS_PROXY env vars
Description:
After upgrading from v2026.4.22 to v2026.4.26, Telegram channel fails to connect
in environments requiring HTTP proxy (e.g., ... | closed | completed | true | 2 | [
"bug"
] | [] | 2026-04-29T02:56:56Z | 2026-05-02T04:28:22Z | 2026-04-29T11:27:37Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | alostlover | 155,522,111 | U_kgDOCUUUPw | User | false |
openclaw/openclaw | 4,347,538,026 | I_kwDOQb6kR88AAAABAyIqag | 74,011 | https://github.com/openclaw/openclaw/issues/74011 | https://api.github.com/repos/openclaw/openclaw/issues/74011 | [Bug]: Slack channel provider crash-loops on Windows with ERR_UNSUPPORTED_ESM_URL_SCHEME ('Received protocol c:') after wizard-driven setup | ### Bug type
Crash (process/app exits or hangs)
### Beta release blocker
No
### Summary
## Bug type
Crash (process/app exits or hangs) — Slack channel provider enters
auto-restart loop and never connects.
## Beta release blocker
No (workaround: disable Slack channel in config)
## Summary
After completing the int... | closed | completed | true | 1 | [
"bug"
] | [] | 2026-04-29T02:53:18Z | 2026-05-01T04:40:46Z | 2026-04-29T02:57:36Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | Stellarjester315 | 254,353,553 | U_kgDODykgkQ | User | false |
openclaw/openclaw | 4,347,551,080 | I_kwDOQb6kR88AAAABAyJdaA | 74,015 | https://github.com/openclaw/openclaw/issues/74015 | https://api.github.com/repos/openclaw/openclaw/issues/74015 | [Feature]: Support writable user prompt modification in existing Plugin SDK hooks | ### Summary
The existing llm_input and llm_output plugin hooks are fire-and-forget (return void). This proposal requests making them mutable — allowing hook handlers to return modified prompt, systemPrompt, historyMessages, and assistantTexts — so plugins can transparently transform content before it reaches the LLM a... | open | null | false | 1 | [
"enhancement"
] | [] | 2026-04-29T02:57:43Z | 2026-05-03T01:38:29Z | null | NONE | null | 20260503T030030Z | 2026-05-03T03:00:30Z | niudaii | 41,844,046 | MDQ6VXNlcjQxODQ0MDQ2 | User | false |
openclaw/openclaw | 4,347,555,157 | I_kwDOQb6kR88AAAABAyJtVQ | 74,017 | https://github.com/openclaw/openclaw/issues/74017 | https://api.github.com/repos/openclaw/openclaw/issues/74017 | oc llm scorer incorrectly marks correct MCQ final answers wrong when ambiguity_flag=true | # oc llm scorer incorrectly marks correct MCQ final answers wrong when ambiguity_flag=true
## Summary
The `oc llm` benchmark scorer appears to reject MCQ answers when `ambiguity_flag=true`, even when the final extracted answer matches `expected_answer`. This causes valid final-answer outputs from native reasoning mod... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T02:59:13Z | 2026-05-01T04:40:32Z | 2026-04-29T03:21:52Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | jmystaki-create | 267,217,950 | U_kgDOD-1sHg | User | false |
openclaw/openclaw | 4,347,555,437 | I_kwDOQb6kR88AAAABAyJubQ | 74,018 | https://github.com/openclaw/openclaw/issues/74018 | https://api.github.com/repos/openclaw/openclaw/issues/74018 | Feature request: Split chat from activity/workbench in Control UI | ## Summary
Add an optional split-view layout to the OpenClaw Control UI: one clean, readable chat column, plus a separate activity/workbench column for tool calls, command output, file edits, progress/status updates, and verification results.
## Problem
Right now, conversational replies and execution details can occup... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T02:59:20Z | 2026-05-01T04:40:45Z | 2026-04-29T03:03:59Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | scott4686 | 17,067,174 | MDQ6VXNlcjE3MDY3MTc0 | User | false |
openclaw/openclaw | 4,347,559,374 | I_kwDOQb6kR88AAAABAyJ9zg | 74,019 | https://github.com/openclaw/openclaw/issues/74019 | https://api.github.com/repos/openclaw/openclaw/issues/74019 | [Bug]: Lobster workflow calling llm-task fails with “No callable tools remain” when tools.alsoAllow includes lobster/llm-task | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Running a .lobster workflow through `/tools/invoke` works until the workflow calls `llm-task`.
`llm-task` appears to start an embedded model run with tools intentionally disabled, but the explicit tool allowlis... | closed | completed | false | 2 | [
"bug"
] | [] | 2026-04-29T03:00:44Z | 2026-05-03T13:42:33Z | 2026-05-03T13:42:33Z | NONE | null | 20260503T150031Z | 2026-05-03T15:00:31Z | technicallysecure | 92,497,344 | U_kgDOBYNlwA | User | false |
openclaw/openclaw | 4,347,559,944 | I_kwDOQb6kR88AAAABAyKACA | 74,020 | https://github.com/openclaw/openclaw/issues/74020 | https://api.github.com/repos/openclaw/openclaw/issues/74020 | Gateway startup: models.mode=replace should skip pricing fetches | When starting the gateway with \models.mode=replace\, the gateway still runs pricing fetches (e.g. OpenRouter, LiteLLM) which take up to 60 seconds to timeout. Since the mode is \
eplace\ and models are manually configured, these fetches are unnecessary and significantly delay startup. Please skip pricing fetches when ... | open | null | false | 1 | [] | [] | 2026-04-29T03:00:58Z | 2026-05-03T10:19:49Z | null | NONE | null | 20260503T120038Z | 2026-05-03T12:00:38Z | CassidyTTWD-bot | 263,262,893 | U_kgDOD7ESrQ | User | false |
openclaw/openclaw | 4,226,978,969 | I_kwDOQb6kR8778pSZ | 63,325 | https://github.com/openclaw/openclaw/issues/63325 | https://api.github.com/repos/openclaw/openclaw/issues/63325 | [Bug]: Gateway SSE streaming corrupts response when stripping <final> tags | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
When using /v1/chat/completions with stream: true, the gateway's <final> tag stripping corrupts the response — it removes actual content lines along with the tags.
### Steps to reproduce
1. Send a streaming r... | closed | completed | true | 1 | [
"bug",
"bug:behavior"
] | [] | 2026-04-08T19:06:42Z | 2026-05-01T04:40:43Z | 2026-04-29T03:13:33Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | tzwickl | 4,968,212 | MDQ6VXNlcjQ5NjgyMTI= | User | false |
openclaw/openclaw | 4,347,598,347 | I_kwDOQb6kR88AAAABAyMWCw | 74,031 | https://github.com/openclaw/openclaw/issues/74031 | https://api.github.com/repos/openclaw/openclaw/issues/74031 | [Feature]: Accessibility: Add adjustable font size for elderly users and accessibility | ### Summary
## Feature Request: Adjustable Font Size
### Problem
Many users, especially elderly people or those with visual impairments, find the current font size in the QClaw UI too small to read comfortably. This prevents them from using the application effectively.
### Proposed Solution
Add a font size adjustmen... | closed | not_planned | true | 1 | [
"enhancement"
] | [] | 2026-04-29T03:11:22Z | 2026-05-01T04:40:42Z | 2026-04-29T03:15:09Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | liujunxing | 10,227,442 | MDQ6VXNlcjEwMjI3NDQy | User | false |
openclaw/openclaw | 4,347,620,691 | I_kwDOQb6kR88AAAABAyNtUw | 74,034 | https://github.com/openclaw/openclaw/issues/74034 | https://api.github.com/repos/openclaw/openclaw/issues/74034 | Heartbeat + auth fallback + message tool creates cascading delivery-mirror feedback loop | ## Summary
When an agent's heartbeat fires while the primary model provider has invalidated auth tokens (all profiles in cooldown), the fallback chain eventually reaches an alternative provider. That provider's LLM sees session context about "empty messages" and uses the `message` tool to send replies directly to the ... | closed | completed | true | 1 | [] | [] | 2026-04-29T03:18:11Z | 2026-05-01T04:40:32Z | 2026-04-29T03:23:46Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | zote | 35,069 | MDQ6VXNlcjM1MDY5 | User | false |
openclaw/openclaw | 4,347,677,085 | I_kwDOQb6kR88AAAABAyRJnQ | 74,039 | https://github.com/openclaw/openclaw/issues/74039 | https://api.github.com/repos/openclaw/openclaw/issues/74039 | [Bug] Gateway crashes with unhandled 'CIAO ANNOUNCEMENT CANCELLED' promise rejection during Bonjour restart | ## Bug Description
Gateway crashes with an unhandled Promise rejection: `CIAO ANNOUNCEMENT CANCELLED` when Bonjour advertiser gets stuck in `announcing` state and is restarted.
## Steps to Reproduce
1. Have Bonjour/mDNS service running
2. Gateway starts normally
3. After some time, Bonjour advertiser gets stuck in `... | closed | completed | true | 1 | [] | [] | 2026-04-29T03:33:08Z | 2026-05-01T04:40:31Z | 2026-04-29T03:38:26Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | jason20260306-cmyk | 274,491,544 | U_kgDOEFxomA | User | false |
openclaw/openclaw | 4,347,696,458 | I_kwDOQb6kR88AAAABAySVSg | 74,042 | https://github.com/openclaw/openclaw/issues/74042 | https://api.github.com/repos/openclaw/openclaw/issues/74042 | [test] gateway memory leak test | test | closed | completed | true | 0 | [] | [] | 2026-04-29T03:38:47Z | 2026-05-01T04:40:30Z | 2026-04-29T03:39:09Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | xiaotong2026 | 268,956,274 | U_kgDOEAfycg | User | false |
openclaw/openclaw | 4,347,697,683 | I_kwDOQb6kR88AAAABAySaEw | 74,043 | https://github.com/openclaw/openclaw/issues/74043 | https://api.github.com/repos/openclaw/openclaw/issues/74043 | Gateway memory leak: RPC becomes unresponsive after ~24 hours | ## Bug Description
Gateway process remains alive but RPC becomes completely unresponsive after running for approximately 24-25 hours. The process must be killed with `kill -9` and restarted by a heal script.
## Environment
- **OS**: WSL2 (Linux 6.6.87.2-microsoft-standard-WSL2, x64)
- **OpenClaw Version**: v2026.4.12... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T03:39:11Z | 2026-05-01T04:40:29Z | 2026-04-29T03:43:28Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | xiaotong2026 | 268,956,274 | U_kgDOEAfycg | User | false |
openclaw/openclaw | 4,347,712,690 | I_kwDOQb6kR88AAAABAyTUsg | 74,047 | https://github.com/openclaw/openclaw/issues/74047 | https://api.github.com/repos/openclaw/openclaw/issues/74047 | Gateway service pinned to old install path after upgrade; newer path breaks /v1/chat/completions reliability | ## Summary
A staged migration to a newer OpenClaw install exposed a split-install / service-path problem:
- a newer **user install** existed at `/home/ubuntu/.npm-global/lib/node_modules/openclaw` (`2026.4.26`)
- the active gateway service was still pinned to the older **system install** at `/usr/lib/node_modules/ope... | open | null | false | 1 | [] | [] | 2026-04-29T03:43:15Z | 2026-05-03T06:50:46Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | nrubioru | 256,029,441 | U_kgDOD0KzAQ | User | false |
openclaw/openclaw | 4,347,706,250 | I_kwDOQb6kR88AAAABAyS7ig | 74,044 | https://github.com/openclaw/openclaw/issues/74044 | https://api.github.com/repos/openclaw/openclaw/issues/74044 | [Bug]: Discord Channel Flapping | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Discord provider on Raspberry Pi arm64 causes gateway to wedge at ~100% CPU after startup on OpenClaw 2026.4.26. Gateway is healthy when channels.discord.enabled=false.
### Steps to reproduce
1. Run OpenClaw 2026.4.26 (be8... | closed | completed | true | 2 | [
"bug"
] | [] | 2026-04-29T03:41:42Z | 2026-05-04T04:35:51Z | 2026-05-01T18:09:12Z | NONE | null | 20260504T060031Z | 2026-05-04T06:00:31Z | SecureCloudProjO | 88,555,872 | MDQ6VXNlcjg4NTU1ODcy | User | false |
openclaw/openclaw | 4,197,665,190 | I_kwDOQb6kR876M0mm | 59,973 | https://github.com/openclaw/openclaw/issues/59973 | https://api.github.com/repos/openclaw/openclaw/issues/59973 | iMessage DM echo: corrupted prefix breaks text-based dedup (v2026.3.31+) | ## Summary
iMessage DM replies echo back to the agent as new inbound messages. The echo cache fails to match because:
1. **Text TTL too short**: `SENT_MESSAGE_TEXT_TTL_MS = 4000` (4s) but echoes arrive 5-10s later
2. **Corrupted prefix**: Echoed text has leading U+FFFD replacement characters that break exact text mat... | closed | completed | true | 1 | [] | [
"vincentkoc"
] | 2026-04-03T00:55:48Z | 2026-05-01T04:40:26Z | 2026-04-29T04:04:22Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | Orionation | 260,579,663 | U_kgDOD4ghTw | User | false |
openclaw/openclaw | 4,347,772,374 | I_kwDOQb6kR88AAAABAyW91g | 74,056 | https://github.com/openclaw/openclaw/issues/74056 | https://api.github.com/repos/openclaw/openclaw/issues/74056 | OpenCode Go provider: deepseek-v4-pro and deepseek-v4-flash missing from model list | ## Description
When configuring OpenCode Go as a provider through OpenClaw's built-in config UI, the model list does not include deepseek-v4-pro or deepseek-v4-flash, despite these models being available on the OpenCode Go API.
## Steps to Reproduce
1. Open OpenClaw config
2. In the model/provider section, select Op... | closed | completed | true | 1 | [] | [] | 2026-04-29T04:00:48Z | 2026-05-01T04:40:22Z | 2026-04-29T04:08:59Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | qwowboyp | 16,829,262 | MDQ6VXNlcjE2ODI5MjYy | User | false |
openclaw/openclaw | 4,347,790,725 | I_kwDOQb6kR88AAAABAyYFhQ | 74,058 | https://github.com/openclaw/openclaw/issues/74058 | https://api.github.com/repos/openclaw/openclaw/issues/74058 | models.list is slow because it does provider startup and dependency installs on the hot path | On OpenClaw `2026.4.23`, `models.list` is taking far too long. In my environment it was often 45s to 2m+.
This does not appear to be Ollama itself: `http://127.0.0.1:11434/api/tags` responded quickly. The slowdown came from OpenClaw doing extra work during model catalog load, including provider plugin loading, bundled... | closed | not_planned | true | 2 | [] | [] | 2026-04-29T04:07:15Z | 2026-05-01T04:40:22Z | 2026-04-29T04:11:19Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | Kru442 | 187,809,917 | U_kgDOCzHAfQ | User | false |
openclaw/openclaw | 4,337,409,818 | I_kwDOQb6kR88AAAABAoefGg | 72,930 | https://github.com/openclaw/openclaw/issues/72930 | https://api.github.com/repos/openclaw/openclaw/issues/72930 | Health check shows Feishu as OFF/not configured when using multi-account format | ## Bug Description
`openclaw status --deep` shows Feishu channel as **OFF · not configured** even when Feishu is fully configured and working (sending/receiving messages via WebSocket).
## Environment
- OpenClaw: 2026.4.24
- OS: macOS 15.4 (arm64, M4 Pro)
- Node: v24.15.0
- Channel: Feishu (WebSocket mode)
## Confi... | closed | completed | true | 2 | [] | [] | 2026-04-27T16:49:28Z | 2026-05-05T04:25:54Z | 2026-05-03T00:19:38Z | NONE | null | 20260505T060034Z | 2026-05-05T06:00:34Z | chouti | 27,688 | MDQ6VXNlcjI3Njg4 | User | false |
openclaw/openclaw | 4,205,285,412 | I_kwDOQb6kR876p5Ak | 60,946 | https://github.com/openclaw/openclaw/issues/60946 | https://api.github.com/repos/openclaw/openclaw/issues/60946 | Feature: Rich tool error messages with fix suggestions | ## Summary
Enhance tool error responses with contextual troubleshooting hints, similar text suggestions, and actionable fix steps.
## Motivation
When tools fail, the current error messages are functional but minimal. Claude Code's approach includes structured "teachable moment" errors that help the model self-correct ... | open | null | false | 3 | [] | [] | 2026-04-04T16:23:09Z | 2026-05-03T13:20:58Z | null | NONE | null | 20260503T150031Z | 2026-05-03T15:00:31Z | ohmanthatsgreat | 269,280,971 | U_kgDOEAzmyw | User | false |
openclaw/openclaw | 4,225,359,321 | I_kwDOQb6kR8772d3Z | 63,195 | https://github.com/openclaw/openclaw/issues/63195 | https://api.github.com/repos/openclaw/openclaw/issues/63195 | Bug: Control UI and TUI sessions intermittently disappear during normal use | # OpenClaw bug report draft: Control UI / TUI session continuity disappears during normal use
## Summary
OpenClaw local sessions intermittently disappear or become inaccessible during normal use.
This affects both:
- Control UI / WebChat
- TUI
The problem is not explained by a single bad local workflow. After isolat... | closed | completed | true | 2 | [] | [] | 2026-04-08T14:38:59Z | 2026-05-05T04:28:48Z | 2026-05-02T05:08:03Z | NONE | null | 20260505T060034Z | 2026-05-05T06:00:34Z | bond260312-cmyk | 270,016,836 | U_kgDOEBghRA | User | false |
openclaw/openclaw | 4,225,365,644 | I_kwDOQb6kR8772faM | 63,196 | https://github.com/openclaw/openclaw/issues/63196 | https://api.github.com/repos/openclaw/openclaw/issues/63196 | [Bug]: cron run history can say finished ok while task audit still reports lost backing session missing | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
`openclaw tasks audit` can report a cron task as `lost` with `backing session missing` even when the cron run history shows that same job run finished successfully.
In this state:
- `openclaw status` shows a t... | open | null | false | 1 | [] | [] | 2026-04-08T14:39:58Z | 2026-05-02T23:57:23Z | null | CONTRIBUTOR | null | 20260503T000641Z | 2026-05-03T00:06:41Z | andyylin | 51,783,311 | MDQ6VXNlcjUxNzgzMzEx | User | false |
openclaw/openclaw | 4,225,515,002 | I_kwDOQb6kR8773D36 | 63,208 | https://github.com/openclaw/openclaw/issues/63208 | https://api.github.com/repos/openclaw/openclaw/issues/63208 | memory-wiki: 支持子目录结构 | ## 问题描述
当前 `memory-wiki` 的 compile 只扫描各目录根目录下的 `.md` 文件，不支持子目录结构。
## 当前行为
vault layout:
```
sources/
OpenClaw配置文件指南/
AGENTS.md
SOUL.md
agent-skills/
skill-xxx.md
```
运行 `openclaw wiki compile`：
- 只识别 `sources/*.md`（根目录）
- 不识别 `sources/OpenClaw配置文件指南/*.md`（子目录）
测试结果：
- 把文件放到 `sources/` 根目录 → 识别并索引
-... | open | null | false | 1 | [] | [] | 2026-04-08T15:03:57Z | 2026-05-03T01:37:15Z | null | NONE | null | 20260503T030030Z | 2026-05-03T03:00:30Z | arlen8411 | 28,994,439 | MDQ6VXNlcjI4OTk0NDM5 | User | false |
openclaw/openclaw | 4,225,930,376 | I_kwDOQb6kR8774pSI | 63,234 | https://github.com/openclaw/openclaw/issues/63234 | https://api.github.com/repos/openclaw/openclaw/issues/63234 | UnraidClaw plugin spams 50+ "invalid config" lines during startup despite valid config | ## Summary
The UnraidClaw plugin spams `invalid config: serverUrl: must have required property 'serverUrl', apiKey: must have required property 'apiKey'` roughly every 300ms during gateway startup, producing 50+ log lines over ~25 seconds before the gateway reaches `ready` state. Once ready, the plugin loads correctly... | open | null | false | 2 | [] | [] | 2026-04-08T16:08:00Z | 2026-05-03T12:30:55Z | null | NONE | null | 20260503T150031Z | 2026-05-03T15:00:31Z | unraidBob | 269,354,639 | U_kgDOEA4Gjw | User | false |
openclaw/openclaw | 4,226,178,542 | I_kwDOQb6kR8775l3u | 63,257 | https://github.com/openclaw/openclaw/issues/63257 | https://api.github.com/repos/openclaw/openclaw/issues/63257 | Windows Gateway Feishu API timeout 30s at startup | ## Bug: Windows Gateway Feishu API timeout 30s at startup
### Problem Description
On Windows, Gateway startup causes Feishu API (open.feishu.cn) to timeout after 30 seconds, but system network is normal.
### Environment
- **OS**: Windows 11
- **Node.js**: v24.14.1
- **OpenClaw Version**: 2026.4.8
- **Feishu SDK**: ... | open | null | false | 3 | [] | [] | 2026-04-08T16:53:55Z | 2026-05-03T12:36:38Z | null | NONE | null | 20260503T150031Z | 2026-05-03T15:00:31Z | feixue1986 | 138,313,016 | U_kgDOCD59OA | User | false |
openclaw/openclaw | 4,226,221,484 | I_kwDOQb6kR8775wWs | 63,261 | https://github.com/openclaw/openclaw/issues/63261 | https://api.github.com/repos/openclaw/openclaw/issues/63261 | [Bug]: Discord owner auth ignores channels.discord.allowFrom, hiding owner-only tools | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
In v2026.4.8, a Discord DM sender allowlisted via `channels.discord.allowFrom` can still resolve as `senderIsOwner=false`, which hides owner-only tools such as `cron` and `gateway`.
### Steps to reproduce
1. ... | closed | completed | true | 1 | [
"bug",
"bug:behavior"
] | [] | 2026-04-08T17:01:49Z | 2026-05-02T04:28:04Z | 2026-04-29T14:21:44Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | kazemnejad | 2,122,102 | MDQ6VXNlcjIxMjIxMDI= | User | false |
openclaw/openclaw | 4,226,267,556 | I_kwDOQb6kR87757mk | 63,266 | https://github.com/openclaw/openclaw/issues/63266 | https://api.github.com/repos/openclaw/openclaw/issues/63266 | Docker to host migration can leave workspace-local and workspace-local/local paths split, breaking local secrets assumptions | After migrating a local OpenClaw setup from Docker-era state to a native host install, I ran into a second path-layout problem beyond the `/home/node/...` absolute-path leakage already reported elsewhere.
## Summary
The migrated state can leave operational files and assumptions split between:
- `~/.openclaw/workspace... | open | null | false | 2 | [] | [] | 2026-04-08T17:09:17Z | 2026-05-02T23:57:25Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | Jmarzab | 196,876,602 | U_kgDOC7wZOg | User | false |
openclaw/openclaw | 4,226,313,043 | I_kwDOQb6kR8776GtT | 63,269 | https://github.com/openclaw/openclaw/issues/63269 | https://api.github.com/repos/openclaw/openclaw/issues/63269 | [Bug]: Mattermost: group/public channel messages not received via WebSocket (regression in 2026.4.8) | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Mattermost bots do not receive or process WebSocket events for public/group channel messages in 2026.4.8. DMs work correctly.
### Steps to reproduce
1. Configure Mattermost channel with `ch... | open | null | false | 2 | [
"bug",
"regression"
] | [] | 2026-04-08T17:15:49Z | 2026-05-03T12:30:58Z | null | NONE | null | 20260503T150031Z | 2026-05-03T15:00:31Z | evotionbe | 188,244,982 | U_kgDOCzhj9g | User | false |
openclaw/openclaw | 4,226,400,227 | I_kwDOQb6kR8776b_j | 63,279 | https://github.com/openclaw/openclaw/issues/63279 | https://api.github.com/repos/openclaw/openclaw/issues/63279 | 2026.4.8: Large-session overflow/compaction timeout can cascade into GatewayDrainingError + subagent announce loss; fallback chain stale until restart | ## Summary
After upgrading to **OpenClaw 2026.4.8**, very large Telegram sessions repeatedly hit context overflow and compaction timeout. In this incident, that appeared to cascade into prolonged gateway draining/task rejection, repeated subagent announce failures, and surprising fallback behavior.
This does **not** ... | open | null | false | 2 | [] | [] | 2026-04-08T17:30:18Z | 2026-05-03T01:37:20Z | null | NONE | null | 20260503T030030Z | 2026-05-03T03:00:30Z | EthanSK | 13,756,744 | MDQ6VXNlcjEzNzU2NzQ0 | User | false |
openclaw/openclaw | 4,226,652,939 | I_kwDOQb6kR8777ZsL | 63,289 | https://github.com/openclaw/openclaw/issues/63289 | https://api.github.com/repos/openclaw/openclaw/issues/63289 | [Bug]:Agent fabricates tool output for factual queries — calendar tool never invoked despite instrumentation | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
I am a non-developer end user running OpenClaw 2026.4.1 in a Lume VM, and that this was discovered when the agent gave me wrong information about my own on-call schedule. I tried to debug with the help of Opus... | open | null | false | 2 | [
"bug",
"bug:behavior"
] | [] | 2026-04-08T18:15:11Z | 2026-05-02T20:25:57Z | null | NONE | null | 20260502T210030Z | 2026-05-02T21:00:30Z | ChasClogston | 36,055,978 | MDQ6VXNlcjM2MDU1OTc4 | User | false |
openclaw/openclaw | 4,226,679,768 | I_kwDOQb6kR8777gPY | 63,295 | https://github.com/openclaw/openclaw/issues/63295 | https://api.github.com/repos/openclaw/openclaw/issues/63295 | Session reset: /new prompt drops sender envelope, blocking personalized greetings | ## Summary
On a bare `/new` (or `/reset`) in a channel like Telegram, the agent's first turn receives the `BARE_SESSION_RESET_PROMPT_BASE` string **without** the `Sender (untrusted metadata)` envelope block. This means the LLM has no way to see who just started the session — even though the channel adapter already pop... | open | null | false | 2 | [] | [] | 2026-04-08T18:20:07Z | 2026-05-02T23:57:26Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | unitypeaceproject | 258,681,123 | U_kgDOD2spIw | User | false |
openclaw/openclaw | 4,226,820,586 | I_kwDOQb6kR8778Cnq | 63,308 | https://github.com/openclaw/openclaw/issues/63308 | https://api.github.com/repos/openclaw/openclaw/issues/63308 | sessions_send schema encourages invalid sessionKey+label calls | ## Bug: `sessions_send` schema encourages invalid calls by exposing `sessionKey` and `label` as independent optionals
### Summary
`sessions_send` rejects requests when both `sessionKey` and `label` are present, but the model-facing tool schema exposes them as separate optional fields with no mutual-exclusion guidance... | open | null | false | 3 | [] | [] | 2026-04-08T18:40:26Z | 2026-05-02T23:12:22Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | ThalynLabs | 38,703,365 | MDQ6VXNlcjM4NzAzMzY1 | User | false |
openclaw/openclaw | 4,227,154,247 | I_kwDOQb6kR8779UFH | 63,335 | https://github.com/openclaw/openclaw/issues/63335 | https://api.github.com/repos/openclaw/openclaw/issues/63335 | [Bug]: Single turn can enter pathological retry churn with same runId, near-identical prompt size, and stuck processing state | ## Summary
A single assistant turn on OpenClaw 2026.4.8 entered pathological retry churn, repeatedly reissuing near-identical requests to the same model/provider before falling back to other candidates. The session stayed in `processing`, `queueDepth=0`, and no user-visible output appeared until a later fallback finall... | open | null | false | 1 | [] | [] | 2026-04-08T19:38:11Z | 2026-05-02T11:36:08Z | null | NONE | null | 20260502T120029Z | 2026-05-02T12:00:29Z | prudkov | 39,012,659 | MDQ6VXNlcjM5MDEyNjU5 | User | false |
openclaw/openclaw | 4,227,295,335 | I_kwDOQb6kR87792hn | 63,343 | https://github.com/openclaw/openclaw/issues/63343 | https://api.github.com/repos/openclaw/openclaw/issues/63343 | Browser bridge: "tab not found" race in ensureTabAvailable when wsUrl lags CDP discovery | ## Bug Report
### Summary
`ensureTabAvailable()` in `server-context.selection.ts` throws `BrowserTabNotFoundError` ("tab not found") due to a race condition where a newly opened tab's `webSocketDebuggerUrl` (`wsUrl`) hasn't populated in CDP's `/json/list` response yet. This is the dominant failure mode for browser to... | open | null | false | 3 | [] | [] | 2026-04-08T20:02:56Z | 2026-05-02T11:36:10Z | null | CONTRIBUTOR | null | 20260502T120029Z | 2026-05-02T12:00:29Z | OpenCodeEngineer | 261,470,075 | U_kgDOD5W3ew | User | false |
openclaw/openclaw | 4,227,689,590 | I_kwDOQb6kR877_Wx2 | 63,371 | https://github.com/openclaw/openclaw/issues/63371 | https://api.github.com/repos/openclaw/openclaw/issues/63371 | [Task] Harden browser/search stack and evaluate Lightpanda | Objective: update/fix the current web research stack (web_search/web_fetch, SearXNG, PinchTab, bird, canvas) and install/evaluate Lightpanda browser before building a cross-tool test matrix.\n\nScope:\n- verify current tool health\n- fix flaky SearXNG / PinchTab / search path issues\n- install Lightpanda browser if sui... | open | null | false | 1 | [] | [] | 2026-04-08T21:13:09Z | 2026-05-02T16:01:33Z | null | NONE | null | 20260502T180029Z | 2026-05-02T18:00:29Z | creationsunitassistant | 261,931,807 | U_kgDOD5zDHw | User | false |
openclaw/openclaw | 4,227,713,742 | I_kwDOQb6kR877_crO | 63,373 | https://github.com/openclaw/openclaw/issues/63373 | https://api.github.com/repos/openclaw/openclaw/issues/63373 | [Bug]: Raw internal transcript content ([TOOL_CALL]) leaked into user-visible Telegram chat after transcript corruption/repair | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Raw internal transcript content ([TOOL_CALL]) leaked into a user-visible Telegram DM after transcript corruption/repair.
### Steps to reproduce
Exact trigger is unknown, but these conditions were present:
1.... | closed | completed | true | 1 | [
"bug",
"bug:behavior"
] | [] | 2026-04-08T21:17:49Z | 2026-05-05T04:26:51Z | 2026-05-02T16:03:47Z | NONE | null | 20260505T060034Z | 2026-05-05T06:00:34Z | 916decker | 6,379,409 | MDQ6VXNlcjYzNzk0MDk= | User | false |
openclaw/openclaw | 4,227,951,175 | I_kwDOQb6kR878AWpH | 63,392 | https://github.com/openclaw/openclaw/issues/63392 | https://api.github.com/repos/openclaw/openclaw/issues/63392 | [Feature]: Per-agent backup, restore, and clone | ### Summary
Add `--agent <id>` support to `openclaw backup` and `openclaw restore` for per-agent backup, restore, and clone operations.
### Problem to solve
`openclaw backup create` and `openclaw restore` operate on the entire instance. In multi-agent setups, if a single agent gets corrupted (bad workspace edit, bro... | open | null | false | 1 | [] | [] | 2026-04-08T22:01:19Z | 2026-05-02T12:53:48Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | saba-ch | 38,084,512 | MDQ6VXNlcjM4MDg0NTEy | User | false |
openclaw/openclaw | 4,228,389,928 | I_kwDOQb6kR878CBwo | 63,423 | https://github.com/openclaw/openclaw/issues/63423 | https://api.github.com/repos/openclaw/openclaw/issues/63423 | [Bug]: Empty catch blocks swallow errors silently, hindering debugging | ### Bug type
Code quality / Developer experience
### Beta release blocker
No
### Summary
Multiple locations in the codebase use empty catch blocks (`} catch {}`) that silently swallow errors. This makes debugging failures difficult and can mask real issues in production.
### Steps to reproduce
1. Open the follow... | closed | completed | true | 1 | [] | [
"sallyom"
] | 2026-04-08T23:34:52Z | 2026-05-04T04:35:13Z | 2026-05-01T22:14:09Z | NONE | null | 20260504T060031Z | 2026-05-04T06:00:31Z | WadydX | 65,117,428 | MDQ6VXNlcjY1MTE3NDI4 | User | false |
openclaw/openclaw | 4,234,530,741 | I_kwDOQb6kR878Zc-1 | 63,905 | https://github.com/openclaw/openclaw/issues/63905 | https://api.github.com/repos/openclaw/openclaw/issues/63905 | bug(slack): inbound attachments in container sandbox fail with placeholder-only turn and fetch error 'invalid onRequestStart method' | # bug(slack): inbound attachments in container sandbox fail with placeholder-only turn and fetch error `invalid onRequestStart method`
## Summary
In a Slack DM running through a container sandbox, inbound file attachments are recognized (`files[]` is present and the agent turn gets a `[Slack file: ...]` placeholder),... | open | null | false | 6 | [] | [] | 2026-04-09T19:43:01Z | 2026-05-02T18:31:18Z | null | NONE | null | 20260502T210030Z | 2026-05-02T21:00:30Z | rodja | 131,391 | MDQ6VXNlcjEzMTM5MQ== | User | false |
openclaw/openclaw | 4,347,852,311 | I_kwDOQb6kR88AAAABAyb2Fw | 74,064 | https://github.com/openclaw/openclaw/issues/74064 | https://api.github.com/repos/openclaw/openclaw/issues/74064 | Dreaming Promotion — Cron-triggered memory entries saturate ranking, blocking real session recalls | ## 🦞 Issue: Dreaming Promotion — Cron-triggered memory entries saturate ranking, blocking real session recalls
**Repository**: openclaw/openclaw
**Version**: 2026.4.26 (be8c246)
---
## Bug Description
The dreaming promotion system (`memory promote`) consistently produces **0 promotions** despite having numerous ... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T04:23:13Z | 2026-05-01T04:40:21Z | 2026-04-29T04:26:51Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | ShallM | 84,125,764 | MDQ6VXNlcjg0MTI1NzY0 | User | false |
openclaw/openclaw | 4,347,841,434 | I_kwDOQb6kR88AAAABAybLmg | 74,061 | https://github.com/openclaw/openclaw/issues/74061 | https://api.github.com/repos/openclaw/openclaw/issues/74061 | [Bug]: [Bug] Telegram integration fails to send local media (Returns "Media failed") via <final>MEDIA:... | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
After updating to the April 26 release, the Telegram integration fails to deliver local media files. When the agent outputs a valid local file path using the `<final>MEDIA:...` tag, the file is not transmitted. Instead, the ... | open | null | false | 2 | [
"bug"
] | [] | 2026-04-29T04:20:37Z | 2026-05-03T03:37:53Z | null | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | averyavi | 37,434,979 | MDQ6VXNlcjM3NDM0OTc5 | User | false |
openclaw/openclaw | 4,347,846,978 | I_kwDOQb6kR88AAAABAybhQg | 74,063 | https://github.com/openclaw/openclaw/issues/74063 | https://api.github.com/repos/openclaw/openclaw/issues/74063 | [Feature]: 建议添加桌面应用程序（全平台）（官方版），如（codex）这种 | ### Summary
建议添加桌面应用程序（全平台）（官方版），如（codex）这种
### Problem to solve
建议添加桌面应用程序（全平台）（官方版），如（codex）这种
### Proposed solution
建议添加桌面应用程序（全平台）（官方版），如（codex）这种
### Alternatives considered
_No response_
### Impact
建议添加桌面应用程序（全平台）（官方版），如（codex）这种
### Evidence/examples
建议添加桌面应用程序（全平台）（官方版），如（codex）这种
### Additional inf... | closed | not_planned | false | 1 | [
"enhancement"
] | [] | 2026-04-29T04:21:49Z | 2026-05-03T09:14:56Z | 2026-05-03T09:14:56Z | NONE | null | 20260503T120038Z | 2026-05-03T12:00:38Z | LING19996 | 79,239,265 | MDQ6VXNlcjc5MjM5MjY1 | User | false |
openclaw/openclaw | 4,073,129,650 | I_kwDOQb6kR87yxway | 45,427 | https://github.com/openclaw/openclaw/issues/45427 | https://api.github.com/repos/openclaw/openclaw/issues/45427 | Feature: Include trusted sender_name in inbound metadata envelope | ## Summary
The current inbound metadata envelope provides `sender_id` (a trusted, platform-injected identifier) but not a resolved `sender_name`. Agents are forced to do their own name resolution — either via a local registry lookup or, incorrectly, by reading conversational context (untrusted).
This leads to a real ... | open | null | false | 3 | [] | [] | 2026-03-13T19:58:14Z | 2026-05-04T00:06:49Z | null | NONE | null | 20260504T030031Z | 2026-05-04T03:00:31Z | scekker | 104,278,659 | U_kgDOBjcqgw | User | false |
openclaw/openclaw | 4,347,644,814 | I_kwDOQb6kR88AAAABAyPLjg | 74,036 | https://github.com/openclaw/openclaw/issues/74036 | https://api.github.com/repos/openclaw/openclaw/issues/74036 | [Bug]: corpus=sessions memory_search returns 0 results for many keywords despite data being indexed | ### Summary
When using `memory_search` with `corpus: "sessions"` on the builtin memory backend, many search queries return 0 results even though:
- The keywords exist in the indexed session data (verified via direct SQL)
- The same keywords return results with `corpus: "memory"` (memory files work fine)
- FTS5 direct ... | closed | completed | true | 1 | [] | [] | 2026-04-29T03:24:32Z | 2026-05-02T04:28:26Z | 2026-04-29T11:16:34Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | jinzhu1991 | 199,324,600 | U_kgDOC-FzuA | User | false |
openclaw/openclaw | 4,347,610,813 | I_kwDOQb6kR88AAAABAyNGvQ | 74,032 | https://github.com/openclaw/openclaw/issues/74032 | https://api.github.com/repos/openclaw/openclaw/issues/74032 | /status inconsistent in Telegram group thread/topic when normal mention replies work` | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
I’m seeing inconsistent `/status` behavior in a Telegram group thread/topic.
## Environment
- OpenClaw `2026.4.22`
- Telegram channel
- group config includes `requireMention: false`
## What works
Normal mention-based activ... | closed | completed | true | 2 | [
"bug"
] | [] | 2026-04-29T03:15:34Z | 2026-05-05T04:27:36Z | 2026-05-02T11:00:49Z | NONE | null | 20260505T060034Z | 2026-05-05T06:00:34Z | WouldenShyp | 266,935,458 | U_kgDOD-kcog | User | false |
openclaw/openclaw | 4,082,555,332 | I_kwDOQb6kR87zVtnE | 48,239 | https://github.com/openclaw/openclaw/issues/48239 | https://api.github.com/repos/openclaw/openclaw/issues/48239 | Telegram exec approvals can surface in the wrong agent/chat despite agentFilter=main | ## Summary
In a multi-agent Telegram setup, exec approval prompts intended for the main agent can surface in the wrong Telegram-routed agent/chat ("费宁2号"), even though Telegram exec approvals were explicitly configured with `agentFilter=["main"]` and `target="dm"`.
This is not just a UX annoyance: it creates approval... | closed | completed | true | 3 | [
"stale"
] | [] | 2026-03-16T13:39:24Z | 2026-05-04T04:37:22Z | 2026-05-01T05:57:24Z | NONE | null | 20260504T060031Z | 2026-05-04T06:00:31Z | ningfeemic-dev | 239,259,264 | U_kgDODkLOgA | User | false |
openclaw/openclaw | 4,097,279,025 | I_kwDOQb6kR870N4Qx | 49,989 | https://github.com/openclaw/openclaw/issues/49989 | https://api.github.com/repos/openclaw/openclaw/issues/49989 | bug: Synology Chat plugin fails to deliver replies when peer ID contains channel prefix (e.g. "synology-chat:4") | ## Description
When a Synology Chat DM session is created with `dmScope: "per-channel-peer"`, the session's `to` field is set to the full peer string `"synology-chat:4"` rather than just the numeric user ID `"4"`.
When the plugin attempts to send a reply, this value is passed to `parseNumericUserId()` in `src/client.... | closed | completed | true | 4 | [
"stale"
] | [] | 2026-03-18T19:28:17Z | 2026-05-04T04:37:21Z | 2026-05-01T05:57:33Z | NONE | null | 20260504T060031Z | 2026-05-04T06:00:31Z | arsinio | 6,880,186 | MDQ6VXNlcjY4ODAxODY= | User | false |
openclaw/openclaw | 4,347,873,948 | I_kwDOQb6kR88AAAABAydKnA | 74,066 | https://github.com/openclaw/openclaw/issues/74066 | https://api.github.com/repos/openclaw/openclaw/issues/74066 | [Bug]: `cron add` / `cron edit` with invalid `--cron` expression logs gateway error and returns `UNAVAILABLE` instead of `INVALID_REQUEST` (raw croner TypeError/RangeError leaks to CLI + log) | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Server-side cron pattern validation is missing in the gateway handlers for `cron.add` and `cron.update`. When a user passes an invalid `--cron` expression to `openclaw cron add` or `openclaw cron edit`, the cro... | closed | completed | true | 1 | [
"bug"
] | [] | 2026-04-29T04:29:05Z | 2026-05-02T04:28:25Z | 2026-04-29T11:16:39Z | CONTRIBUTOR | null | 20260502T060142Z | 2026-05-02T06:01:42Z | bittoby | 218,712,309 | U_kgDODQlI9Q | User | false |
openclaw/openclaw | 4,099,291,678 | I_kwDOQb6kR870Vjoe | 50,220 | https://github.com/openclaw/openclaw/issues/50220 | https://api.github.com/repos/openclaw/openclaw/issues/50220 | macOS node host: resolveCanonicalApprovalCwdSync fails for /Users/<user> despite valid path | ## Environment
- OpenClaw version: 2026.3.13
- macOS (node host): headless node via launchd
- Gateway: Linux VPS (Ubuntu 22.04)
- Node binary: /opt/homebrew/Cellar/node@22/22.22.0_1/bin/node (FDA granted, TCC auth_value=2)
## Problem
When using `host=node` exec with `workdir=/Users/mattg` (or any path under /Users/), ... | closed | completed | true | 3 | [
"stale"
] | [] | 2026-03-19T04:58:01Z | 2026-05-04T04:37:17Z | 2026-05-01T07:37:30Z | NONE | null | 20260504T060031Z | 2026-05-04T06:00:31Z | pmgreenjr | 26,364,594 | MDQ6VXNlcjI2MzY0NTk0 | User | false |
openclaw/openclaw | 4,103,628,407 | I_kwDOQb6kR870mGZ3 | 50,590 | https://github.com/openclaw/openclaw/issues/50590 | https://api.github.com/repos/openclaw/openclaw/issues/50590 | Sandbox resolves skill SKILL.md to global path instead of sandbox-local copy | ## Summary
When running in sandbox mode (`agents.defaults.sandbox.mode: "non-main"`), OpenClaw resolves skill paths to `~/.openclaw/skills/<name>/SKILL.md` (the global location) instead of the sandbox-local copy. The sandbox boundary blocks reading the global path, causing the skill to fail.
## Steps to Reproduce
1.... | closed | completed | false | 5 | [] | [
"sallyom"
] | 2026-03-19T18:59:49Z | 2026-05-05T05:13:23Z | 2026-05-05T05:13:23Z | NONE | null | 20260505T060034Z | 2026-05-05T06:00:34Z | thunter009 | 7,088,252 | MDQ6VXNlcjcwODgyNTI= | User | false |
openclaw/openclaw | 4,074,464,266 | I_kwDOQb6kR87y22QK | 45,679 | https://github.com/openclaw/openclaw/issues/45679 | https://api.github.com/repos/openclaw/openclaw/issues/45679 | [Feature] Add config option to auto-silent verbose/streaming messages on Telegram | ## Summary
Add a config option to automatically send verbose/streaming messages silently (with `disable_notification: true`) while keeping normal messages with notifications.
## Use Case
When verbose mode is enabled, the agent sends many streaming/verbose messages (thinking process, tool calls, etc.). These are useful... | closed | completed | true | 4 | [
"stale"
] | [] | 2026-03-14T03:04:10Z | 2026-05-02T04:29:35Z | 2026-04-29T04:39:19Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | log-li | 22,101,261 | MDQ6VXNlcjIyMTAxMjYx | User | false |
openclaw/openclaw | 4,347,764,512 | I_kwDOQb6kR88AAAABAyWfIA | 74,055 | https://github.com/openclaw/openclaw/issues/74055 | https://api.github.com/repos/openclaw/openclaw/issues/74055 | [Bug]: v2026.4.26: OAuth Refresh Lock Fails in Multi-Agent Swarm (401 refresh_token_reused) | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Description
In version 2026.4.26, the internal OAuth refresh lock fails to serialize token refresh requests when running a multi-agent swarm sharing a single ChatGPT Plus account (openai-codex provider). Even when strictly ... | closed | completed | true | 4 | [
"bug",
"regression"
] | [] | 2026-04-29T03:58:25Z | 2026-05-02T04:28:19Z | 2026-04-29T12:05:57Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | ClarityInvest | 155,584,766 | U_kgDOCUYI_g | User | false |
openclaw/openclaw | 4,347,917,093 | I_kwDOQb6kR88AAAABAyfzJQ | 74,071 | https://github.com/openclaw/openclaw/issues/74071 | https://api.github.com/repos/openclaw/openclaw/issues/74071 | [Bug] isAnnounceSkip strict equality misses multi-line agent outputs ending with ANNOUNCE_SKIP | **Version:** OpenClaw 2026.4.26 (also reproduced in 2026.4.21 / 2026.4.15)
**File:** `dist/subagent-announce-output-V8Qd-Aa3.js` (hash floats; the file containing `function isAnnounceSkip`)
**Symptom**
Sub-agents that produce a multi-line summary followed by `ANNOUNCE_SKIP` on the final line have their entire summar... | open | null | false | 1 | [] | [] | 2026-04-29T04:38:16Z | 2026-05-03T01:38:32Z | null | NONE | null | 20260503T030030Z | 2026-05-03T03:00:30Z | dcapclaw | 265,159,728 | U_kgDOD84EMA | User | false |
openclaw/openclaw | 4,347,917,091 | I_kwDOQb6kR88AAAABAyfzIw | 74,070 | https://github.com/openclaw/openclaw/issues/74070 | https://api.github.com/repos/openclaw/openclaw/issues/74070 | [Bug] resolveBackgroundTaskTerminalResult returns empty terminalSummary; progressSummary truncated at 240 chars | **Version:** OpenClaw 2026.4.26 (also 4.21 / 4.15)
**File:** `dist/manager-BIQ5eGBZ.js` — `resolveBackgroundTaskTerminalResult()`
**Symptom**
Background-task completions delivered via `sessions_send` arrive as bare "Background task done" with no payload. The `terminalSummary` field is `null` for non-error completion... | open | null | false | 1 | [] | [] | 2026-04-29T04:38:16Z | 2026-05-03T12:31:56Z | null | NONE | null | 20260503T150031Z | 2026-05-03T15:00:31Z | dcapclaw | 265,159,728 | U_kgDOD84EMA | User | false |
openclaw/openclaw | 4,347,918,103 | I_kwDOQb6kR88AAAABAyf3Fw | 74,074 | https://github.com/openclaw/openclaw/issues/74074 | https://api.github.com/repos/openclaw/openclaw/issues/74074 | ACP task maintenance accepts stale session-store rows as valid backing sessions | ACP task maintenance trusts `session-store` row existence too blindly. When a stale/invalid session-store entry exists for an ACP task, maintenance incorrectly treats it as a live backing session, preventing proper reconciliation and "lost" marking.
## Observed Behavior
1. ACP run leaves a stale session-store row (e.... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T04:38:30Z | 2026-05-02T04:29:31Z | 2026-04-29T04:42:31Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | WindoC | 10,722,906 | MDQ6VXNlcjEwNzIyOTA2 | User | false |
openclaw/openclaw | 4,347,917,096 | I_kwDOQb6kR88AAAABAyfzKA | 74,072 | https://github.com/openclaw/openclaw/issues/74072 | https://api.github.com/repos/openclaw/openclaw/issues/74072 | [Bug] BOOT.md fires per-agent on shared workspaces — duplicate startup announcements on gateway restart | **Version:** OpenClaw 2026.4.26 (long-standing — reproduced from 2026.4.x)
**File:** `dist/bundled/boot-md/handler.js` — `runBootChecklist`
**Symptom**
When multiple agents share the same workspace directory (e.g. a primary orchestrator agent + several specialist agents all rooted at `~/Jarvis-V2`), `BOOT.md` fires ... | closed | completed | true | 1 | [] | [] | 2026-04-29T04:38:16Z | 2026-05-02T04:28:36Z | 2026-04-29T09:48:33Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | dcapclaw | 265,159,728 | U_kgDOD84EMA | User | false |
openclaw/openclaw | 4,347,917,097 | I_kwDOQb6kR88AAAABAyfzKQ | 74,073 | https://github.com/openclaw/openclaw/issues/74073 | https://api.github.com/repos/openclaw/openclaw/issues/74073 | [Bug] Auto-compaction race: readSubagentOutput returns stale tool JSON instead of post-compaction assistant reply | **Version:** OpenClaw 2026.4.26 (reproduced from 2026.4.15 onward)
**File:** `dist/subagent-announce-Dim2fAZI.js` (hash floats; the file with `runSubagentAnnounceFlow` / `childCompletionFindings`)
**Symptom**
When a sub-agent uses tools that trigger auto-compaction mid-turn (most reliably reproduced with Grok / Deep... | open | null | false | 2 | [] | [] | 2026-04-29T04:38:16Z | 2026-05-03T09:13:29Z | null | NONE | null | 20260503T120038Z | 2026-05-03T12:00:38Z | dcapclaw | 265,159,728 | U_kgDOD84EMA | User | false |
openclaw/openclaw | 4,347,897,130 | I_kwDOQb6kR88AAAABAyelKg | 74,069 | https://github.com/openclaw/openclaw/issues/74069 | https://api.github.com/repos/openclaw/openclaw/issues/74069 | [Bug]: 426 can not downgrade | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
426 can not downgrade to 425 version
### Steps to reproduce
:~/.openclaw$ openclaw update --tag 2026.4.25
│
◇ Downgrading from 2026.4.26 to 2026.4.25 can break configuration. Continue?
│ Yes
Updating OpenClaw...
│
◇ ✓ ... | open | null | false | 1 | [
"bug"
] | [] | 2026-04-29T04:34:08Z | 2026-05-03T06:50:49Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | sakalaboator | 42,799,107 | MDQ6VXNlcjQyNzk5MTA3 | User | false |
openclaw/openclaw | 4,347,927,378 | I_kwDOQb6kR88AAAABAygbUg | 74,076 | https://github.com/openclaw/openclaw/issues/74076 | https://api.github.com/repos/openclaw/openclaw/issues/74076 | [Bug] Windows: deepseek-v4-flash 400 error + Web UI CPU/memory high + openclaw-weixin channelConfigs warning (v2026.4.22→v2026.4.26) | ## Summary
Windows 11 上 OpenClaw v2026.4.22 → v2026.4.26 升级后，deepseek-v4-flash 模型多轮对话 API 报错、Web UI 性能严重退化、微信插件 openclaw-weixin 持续报配置警告。**这三个问题相互关联，均在使用 deepseek-v4-flash 模型时暴露。**
---
## Issue 1: deepseek-v4-flash reasoning_content 400 错误
**现象：** 多轮对话后 API 返回 400 错误。
```
400 The `reasoning_content` in the thinking... | closed | not_planned | true | 1 | [] | [] | 2026-04-29T04:40:19Z | 2026-05-02T04:29:29Z | 2026-04-29T04:44:44Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | caiyc16888 | 266,687,462 | U_kgDOD-VT5g | User | false |
openclaw/openclaw | 4,347,939,218 | I_kwDOQb6kR88AAAABAyhJkg | 74,077 | https://github.com/openclaw/openclaw/issues/74077 | https://api.github.com/repos/openclaw/openclaw/issues/74077 | Feature: slash command to set preview streaming mode for current chat/session | ## Feature request
Add a slash/native command to change preview streaming behavior for the current chat/session without editing gateway config or restarting the gateway.
## Proposed UX
```text
/stream
/stream off
/stream final
/stream partial
/stream progress
/stream reset
```
`/stream` with no args would show the ... | open | null | false | 1 | [] | [] | 2026-04-29T04:42:38Z | 2026-05-03T06:50:50Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | clawSean | 260,045,960 | U_kgDOD3_8iA | User | false |
openclaw/openclaw | 4,347,967,636 | I_kwDOQb6kR88AAAABAyi4lA | 74,079 | https://github.com/openclaw/openclaw/issues/74079 | https://api.github.com/repos/openclaw/openclaw/issues/74079 | [Bug] Billing-error classifier misses snake_case top-level `{"error":"insufficient_balance",…}` payloads — raw JSON leaks to user-facing chat | ## Summary
When an LLM provider/proxy returns a 402-style billing block as a JSON payload **without using the OpenAI/Anthropic-style envelope**, openclaw's `formatAssistantErrorText` falls through every classifier and emits the raw JSON unchanged into the assistant's user-facing reply.
In production this surfaces to ... | closed | completed | true | 1 | [] | [] | 2026-04-29T04:49:29Z | 2026-05-02T04:28:26Z | 2026-04-29T11:15:47Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | jarvis-ai-gregmoser | 258,473,251 | U_kgDOD2f9Iw | User | false |
openclaw/openclaw | 4,347,761,091 | I_kwDOQb6kR88AAAABAyWRww | 74,054 | https://github.com/openclaw/openclaw/issues/74054 | https://api.github.com/repos/openclaw/openclaw/issues/74054 | [Bug]: active-memory plugin infinite retry on API timeout causes token waste | ## Bug type
Behavior bug (incorrect output/state without crash)
## Beta release blocker
No
## Summary
When using `xiaomi/mimo-v2-flash` model with the `active-memory` plugin, API calls frequently timeout (15s threshold), but the plugin continues to retry indefinitely without any retry limit configuration. This cau... | closed | completed | true | 2 | [] | [] | 2026-04-29T03:57:07Z | 2026-05-02T04:28:24Z | 2026-04-29T11:18:45Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | ariesy | 956,025 | MDQ6VXNlcjk1NjAyNQ== | User | false |
openclaw/openclaw | 4,347,954,374 | I_kwDOQb6kR88AAAABAyiExg | 74,078 | https://github.com/openclaw/openclaw/issues/74078 | https://api.github.com/repos/openclaw/openclaw/issues/74078 | [Bug]: 4.26 version cutdown connection to local model in few minites | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
After upgrade to 426. The task will hang and the connection to LLM local model seems be cut off.
### Steps to reproduce
The task will hang down and the connection to LLM local model(ollama) seems be cut off. The same conf... | closed | completed | true | 4 | [
"bug"
] | [] | 2026-04-29T04:46:09Z | 2026-05-02T04:26:33Z | 2026-04-30T02:45:24Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | sakalaboator | 42,799,107 | MDQ6VXNlcjQyNzk5MTA3 | User | false |
openclaw/openclaw | 4,185,745,477 | I_kwDOQb6kR875fWhF | 58,932 | https://github.com/openclaw/openclaw/issues/58932 | https://api.github.com/repos/openclaw/openclaw/issues/58932 | [Bug]: normalizeHyphenSlug strips all CJK characters, breaking group display names for non-Latin languages | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
`normalizeHyphenSlug` regex `/[^a-z0-9#@._+-]+/g` strips all non-ASCII characters, causing `buildGroupDisplayName` to return only the provider key for group names containing CJK/Cyrillic/Arabic characters.
##... | closed | completed | true | 2 | [
"bug"
] | [] | 2026-04-01T10:09:02Z | 2026-05-02T04:29:27Z | 2026-04-29T04:56:58Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | fengqing-git | 93,363,010 | U_kgDOBZCbQg | User | false |
openclaw/openclaw | 4,347,984,492 | I_kwDOQb6kR88AAAABAyj6bA | 74,080 | https://github.com/openclaw/openclaw/issues/74080 | https://api.github.com/repos/openclaw/openclaw/issues/74080 | [Bug]: Subagent completion announce uses child agent's accountId instead of parent session's delivery context, causing Feishu "open_id cross app" 400 error | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Description
When a parent agent (bound to Feishu app A) spawns a subagent that runs under a different agent (bound to Feishu app B), the subagent completion announce delivery incorrectly uses the child agent's accountId (app... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-29T04:53:23Z | 2026-05-02T04:29:26Z | 2026-04-29T04:58:03Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | zdh1207Git | 31,314,029 | MDQ6VXNlcjMxMzE0MDI5 | User | false |
openclaw/openclaw | 4,065,118,437 | I_kwDOQb6kR87yTMjl | 44,130 | https://github.com/openclaw/openclaw/issues/44130 | https://api.github.com/repos/openclaw/openclaw/issues/44130 | [TUI] Scroll-jump / auto-scroll behavior is still disruptive in 2026.3.8 | ## Summary
The TUI still has a scroll-jump problem during normal chat use.
When I send a new message/order in `openclaw tui`, the view jumps/scrolls in a very disruptive way instead of staying stable where I am reading. In practice it feels like the screen scrolls from top to end and makes the conversation annoying t... | open | null | false | 2 | [] | [] | 2026-03-12T14:34:48Z | 2026-05-04T12:46:08Z | null | NONE | null | 20260504T180039Z | 2026-05-04T18:00:39Z | WingsOfPanda | 46,889,518 | MDQ6VXNlcjQ2ODg5NTE4 | User | false |
openclaw/openclaw | 4,348,009,583 | I_kwDOQb6kR88AAAABAylcbw | 74,085 | https://github.com/openclaw/openclaw/issues/74085 | https://api.github.com/repos/openclaw/openclaw/issues/74085 | Bug: openclaw status --usage --json hangs/fails from non-TTY subprocess in 2026.4.26 | After upgrading from 2026.4.23 to 2026.4.26, `openclaw status --usage --json` no longer works reliably when called from a non-TTY subprocess. It works from an interactive terminal, but hangs or fails with empty stdout/stderr when invoked from automation.
## Environment
- macOS arm64
- OpenClaw 2026.4.26
- Node v22.22... | open | null | false | 1 | [] | [] | 2026-04-29T04:59:27Z | 2026-05-03T07:45:34Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | DoubleAces | 417,534 | MDQ6VXNlcjQxNzUzNA== | User | false |
openclaw/openclaw | 4,065,301,734 | I_kwDOQb6kR87yT5Tm | 44,148 | https://github.com/openclaw/openclaw/issues/44148 | https://api.github.com/repos/openclaw/openclaw/issues/44148 | [Bug]:WhatsApp: Pause bot when human operator replies manually (outgoing message detection | ### Bug type
Regression (worked before, now fails)
### Summary
Description:
When running Clawdbot on a WhatsApp number used by a retail store, human operators sometimes reply to customers manually via WhatsApp Web on the same number the bot is running on.
The bot does not detect these outgoing messages and keeps re... | closed | not_planned | true | 2 | [
"bug",
"regression"
] | [] | 2026-03-12T15:03:16Z | 2026-05-02T04:28:10Z | 2026-04-29T13:43:42Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | dimafer-gif | 246,135,540 | U_kgDODqu69A | User | false |
openclaw/openclaw | 4,065,388,815 | I_kwDOQb6kR87yUOkP | 44,160 | https://github.com/openclaw/openclaw/issues/44160 | https://api.github.com/repos/openclaw/openclaw/issues/44160 | Mattermost WebSocket: add ping/pong keepalive to detect silent connection drops | ## Problem
The Mattermost plugin's WebSocket connection has no ping/pong heartbeat mechanism. When the TCP connection dies silently (e.g. NAT table expiry, intermediate router drops), no `close` or `error` event fires on the WebSocket. The `runWithReconnect` loop never triggers because it depends on these events.
**R... | closed | not_planned | true | 2 | [] | [] | 2026-03-12T15:16:16Z | 2026-05-02T04:27:58Z | 2026-04-29T15:03:09Z | NONE | null | 20260502T060142Z | 2026-05-02T06:01:42Z | allenliang2022 | 98,301,349 | U_kgDOBdv1pQ | User | false |
openclaw/openclaw | 4,065,524,485 | I_kwDOQb6kR87yUvsF | 44,179 | https://github.com/openclaw/openclaw/issues/44179 | https://api.github.com/repos/openclaw/openclaw/issues/44179 | GLM-5 token leak: stripModelSpecialTokens not applied to main output path | ## Problem
OpenClaw 3.11 implemented a fix for GLM-5 token leakage (`stripModelSpecialTokens`), but the fix is not applied to the main chat output path, so users still see token leakage in their conversations.
## Analysis
### ✅ Fix Implemented
The fix exists in `./src/agents/pi-embedded-utils.ts`:
\`\`\`typescript... | open | null | false | 3 | [] | [] | 2026-03-12T15:36:00Z | 2026-05-04T12:46:27Z | null | NONE | null | 20260504T180039Z | 2026-05-04T18:00:39Z | king33yoyo | 18,387,267 | MDQ6VXNlcjE4Mzg3MjY3 | User | false |
openclaw/openclaw | 4,066,354,218 | I_kwDOQb6kR87yX6Qq | 44,289 | https://github.com/openclaw/openclaw/issues/44289 | https://api.github.com/repos/openclaw/openclaw/issues/44289 | Generate secretref reference docs from secret target registry metadata | ## Problem
`src/secrets/target-registry.test.ts` currently catches drift between code and docs after the fact, but the secretref reference docs are still manually synced:
- `docs/reference/secretref-user-supplied-credentials-matrix.json`
- `docs/reference/secretref-credential-surface.md`
This creates recurring branc... | open | null | false | 2 | [] | [] | 2026-03-12T18:05:43Z | 2026-05-04T12:46:43Z | null | NONE | null | 20260504T180039Z | 2026-05-04T18:00:39Z | aniruddhaadak80 | 127,435,065 | U_kgDOB5iBOQ | User | false |
openclaw/openclaw | 4,066,354,354 | I_kwDOQb6kR87yX6Sy | 44,291 | https://github.com/openclaw/openclaw/issues/44291 | https://api.github.com/repos/openclaw/openclaw/issues/44291 | Add native PowerShell smoke coverage for contributor commands | ## Problem
Several contributor-facing commands are effectively validated only through bash-style environments. Recent examples included commands that worked in Git Bash/CI but failed in native PowerShell due to shell assumptions.
## Why It Matters
- Native Windows contributors discover shell-specific failures late.
... | open | null | false | 2 | [] | [] | 2026-03-12T18:05:45Z | 2026-05-04T12:46:45Z | null | NONE | null | 20260504T180039Z | 2026-05-04T18:00:39Z | aniruddhaadak80 | 127,435,065 | U_kgDOB5iBOQ | User | false |
openclaw/openclaw | 4,066,354,466 | I_kwDOQb6kR87yX6Ui | 44,292 | https://github.com/openclaw/openclaw/issues/44292 | https://api.github.com/repos/openclaw/openclaw/issues/44292 | Add an autofix or generator for missing config field labels | ## Problem
`src/config/schema.help.quality.test.ts` fails when new help keys are added without corresponding entries in `FIELD_LABELS`. A recent example was `gateway.push`, which caused CI failures until labels were added manually.
## Why It Matters
- The current workflow is easy to forget during unrelated feature w... | open | null | false | 2 | [] | [] | 2026-03-12T18:05:46Z | 2026-05-04T12:46:46Z | null | NONE | null | 20260504T180039Z | 2026-05-04T18:00:39Z | aniruddhaadak80 | 127,435,065 | U_kgDOB5iBOQ | User | false |
openclaw/openclaw | 4,066,354,539 | I_kwDOQb6kR87yX6Vr | 44,293 | https://github.com/openclaw/openclaw/issues/44293 | https://api.github.com/repos/openclaw/openclaw/issues/44293 | Make `pnpm check:docs` work in native PowerShell | ## Problem
`pnpm check:docs` currently depends on a Unix-style pipeline in `format:docs:check`:
```sh
git ls-files 'docs/**/*.md' 'docs/**/*.mdx' 'README.md' | xargs oxfmt --check
```
That works in bash/Git Bash, but fails in native Windows PowerShell because `xargs` is not available by default.
## Why It Matters
... | open | null | false | 2 | [] | [] | 2026-03-12T18:05:47Z | 2026-05-04T12:46:48Z | null | NONE | null | 20260504T180039Z | 2026-05-04T18:00:39Z | aniruddhaadak80 | 127,435,065 | U_kgDOB5iBOQ | User | false |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.