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 | 4,040,126,104 | I_kwDOQb6kR87wz26Y | 39,343 | https://github.com/openclaw/openclaw/issues/39343 | https://api.github.com/repos/openclaw/openclaw/issues/39343 | Feature Request: Image batching / media group buffering at gateway layer | ## Problem
When a user sends multiple images in rapid succession (e.g. 5 photos in LINE or a Telegram album), each image arrives as a separate update and gets forwarded to the agent individually. This causes the agent to either:
1. Reply to each image separately (spamming the chat with N responses)
2. Implement awkwa... | open | null | false | 3 | [] | [] | 2026-03-08T01:27:54Z | 2026-05-04T08:06:44Z | null | NONE | null | 20260504T090105Z | 2026-05-04T09:01:05Z | walkpod1007 | 258,314,818 | U_kgDOD2WSQg | User | false |
openclaw/openclaw | 4,040,177,431 | I_kwDOQb6kR87w0DcX | 39,359 | https://github.com/openclaw/openclaw/issues/39359 | https://api.github.com/repos/openclaw/openclaw/issues/39359 | [Bug]: Cron job with `sessionTarget: "isolated"` fails on tool schema validation: "Field required" on `tools.function` | ### Bug type
Behavior bug (incorrect output/state without crash)
### Summary
**Environment**:
- OpenClaw version: 2026.2.13 (203b5bd)
- Node.js: v22.22.0
- OS: Ubuntu 24.04 LTS (GNU/Linux 6.8.0-55-generic x86_64)
- Deployment: Gateway (systemd)
**Description**:
When running a cron job with `sessionTarget: "isolated... | closed | completed | true | 5 | [
"bug",
"stale",
"bug:behavior"
] | [] | 2026-03-08T01:51:59Z | 2026-04-27T04:59:55Z | 2026-04-24T20:30:41Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | effymengxue | 10,512,357 | MDQ6VXNlcjEwNTEyMzU3 | User | false |
openclaw/openclaw | 4,040,185,030 | I_kwDOQb6kR87w0FTG | 39,362 | https://github.com/openclaw/openclaw/issues/39362 | https://api.github.com/repos/openclaw/openclaw/issues/39362 | Internal hook message:sent fires inconsistently for group/topic conversations | ## Summary
The `message:sent` internal hook fires inconsistently. It registered correctly and fired once during a gateway restart, but does not fire on subsequent outbound messages to Telegram group topics.
## Steps to Reproduce
1. Create an internal hook in `~/.openclaw/hooks/table-guard/` listening on `message:sen... | closed | completed | true | 3 | [
"stale"
] | [] | 2026-03-08T01:56:20Z | 2026-04-27T04:59:55Z | 2026-04-24T20:30:44Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | dobbysss | 261,629,008 | U_kgDOD5gkUA | User | false |
openclaw/openclaw | 4,040,204,617 | I_kwDOQb6kR87w0KFJ | 39,373 | https://github.com/openclaw/openclaw/issues/39373 | https://api.github.com/repos/openclaw/openclaw/issues/39373 | [Feature]: Web UI 组件库现代化升级 | ## 背景
当前 OpenClaw Web UI 组件库使用的是基础样式，建议进行现代化升级以提升用户体验和开发效率。
## 目标
1. 评估当前 Web UI 组件现状
2. 调研现代化 UI 框架（如 Tailwind CSS、shadcn/ui 等）
3. 制定组件库升级路线图
## 可见结果
- [ ] 产出当前组件库现状分析报告
- [ ] 产出技术选型对比文档 | open | null | false | 3 | [] | [] | 2026-03-08T02:10:47Z | 2026-05-04T07:01:06Z | null | NONE | null | 20260504T090105Z | 2026-05-04T09:01:05Z | liangruochong44-ui | 229,361,665 | U_kgDODavIAQ | User | false |
openclaw/openclaw | 4,040,214,807 | I_kwDOQb6kR87w0MkX | 39,380 | https://github.com/openclaw/openclaw/issues/39380 | https://api.github.com/repos/openclaw/openclaw/issues/39380 | Control UI: 'Unsupported schema node' error in Agent List | ## Bug Description
Control UI shows "Unsupported schema node" error when viewing/editing Agent configurations in the Agents section. Also displays "Form view can't safely edit some fields. Use Raw to avoid losing config entries."
## Steps to Reproduce
1. Open Control UI (web interface)
2. Navigate to Agents section
3.... | closed | completed | true | 3 | [
"stale"
] | [] | 2026-03-08T02:17:26Z | 2026-04-27T04:59:54Z | 2026-04-24T20:30:52Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | aplake-lsc | 185,314,373 | U_kgDOCwusRQ | User | false |
openclaw/openclaw | 4,040,217,701 | I_kwDOQb6kR87w0NRl | 39,382 | https://github.com/openclaw/openclaw/issues/39382 | https://api.github.com/repos/openclaw/openclaw/issues/39382 | Feishu binding without explicit accountId silently falls back to main | # Issue: Feishu binding without explicit `accountId` silently falls back to main
## Summary
Feishu peer bindings can appear correctly configured but still silently route to the default/main agent if `match.accountId` is omitted.
In current OpenClaw behavior, an empty `match.accountId` does **not** mean wildcard. Ins... | closed | completed | true | 2 | [
"stale"
] | [] | 2026-03-08T02:19:41Z | 2026-04-28T04:43:10Z | 2026-04-26T02:00:23Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Meskjei | 24,961,623 | MDQ6VXNlcjI0OTYxNjIz | User | false |
openclaw/openclaw | 4,040,239,436 | I_kwDOQb6kR87w0SlM | 39,392 | https://github.com/openclaw/openclaw/issues/39392 | https://api.github.com/repos/openclaw/openclaw/issues/39392 | [Bug]: Agent switch causes stale/wrong model selection UI in Control UI | ### Bug type
Behavior bug (incorrect output/state without crash)
### Summary
Switching from main → youtubeagent → main can make the model dropdown in Control UI show the wrong value for main, even though runtime/config remain correct.
### Steps to reproduce
1. Open Control UI → Agents.
2. Select `main` and confirm... | closed | completed | true | 5 | [
"bug",
"bug:behavior"
] | [] | 2026-03-08T02:30:58Z | 2026-04-30T04:33:36Z | 2026-04-27T19:06:04Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | daijunyi002 | 215,509,586 | U_kgDODNhqUg | User | false |
openclaw/openclaw | 4,040,265,405 | I_kwDOQb6kR87w0Y69 | 39,400 | https://github.com/openclaw/openclaw/issues/39400 | https://api.github.com/repos/openclaw/openclaw/issues/39400 | session_status usage data is unstable: Day window jumps and usage fields intermittently missing | ### Summary
When running a cron-based usage monitor (built on `session_status`), Day window usage appears unstable:
1. **Day percentage/reset values sometimes jump unexpectedly** (e.g., after a normal downward trend, suddenly returning near/at 100%).
2. **Usage fields are intermittently missing** — some runs return on... | open | null | false | 2 | [] | [] | 2026-03-08T02:49:17Z | 2026-05-04T07:01:10Z | null | NONE | null | 20260504T090105Z | 2026-05-04T09:01:05Z | hualinsen | 3,954,604 | MDQ6VXNlcjM5NTQ2MDQ= | User | false |
openclaw/openclaw | 4,040,278,873 | I_kwDOQb6kR87w0cNZ | 39,406 | https://github.com/openclaw/openclaw/issues/39406 | https://api.github.com/repos/openclaw/openclaw/issues/39406 | Feature request: config option to suppress transient tool error warnings | ## Summary
When a tool fails transiently (e.g., an `edit` with mismatched content) but the agent retries and succeeds, the tool error warning is still delivered to the user's channel as a visible message:
```
⚠️ 📝 Edit: in ~/IDENTITY.md (521 chars) failed
```
This creates a confusing UX where the user sees:
1. "Job... | open | null | false | 2 | [] | [] | 2026-03-08T02:59:19Z | 2026-05-04T07:01:12Z | null | NONE | null | 20260504T090105Z | 2026-05-04T09:01:05Z | neoclaw-latrobe | 264,294,965 | U_kgDOD8DSNQ | User | false |
openclaw/openclaw | 4,040,289,559 | I_kwDOQb6kR87w0e0X | 39,408 | https://github.com/openclaw/openclaw/issues/39408 | https://api.github.com/repos/openclaw/openclaw/issues/39408 | [Bug] Feishu channel: /new startup greeting not delivered to chat | ## Summary
After running `/new` (or `/reset`) in a Feishu DM, the agent generates a startup greeting but it is **never delivered** to the Feishu chat. The user only sees the greeting after sending a follow-up message.
## Environment
- **OpenClaw version:** 2026.3.2
- **Feishu plugin:** `@larksuiteoapi/feishu-opencla... | closed | completed | true | 3 | [
"stale"
] | [] | 2026-03-08T03:06:00Z | 2026-04-27T04:59:52Z | 2026-04-24T20:30:57Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | TeigenZhang | 38,639,187 | MDQ6VXNlcjM4NjM5MTg3 | User | false |
openclaw/openclaw | 4,040,321,723 | I_kwDOQb6kR87w0mq7 | 39,417 | https://github.com/openclaw/openclaw/issues/39417 | https://api.github.com/repos/openclaw/openclaw/issues/39417 | Control UI signs device auth with a different token than it sends, causing device signature invalid | ## Summary
The browser Control UI client can sign the device-auth payload with a different token than the one it actually sends in the `connect` auth object. When that happens, the gateway rejects the connection with `device signature invalid`.
This appears to be a client bug, not a deployment/config issue.
## Serve... | closed | completed | true | 4 | [
"stale"
] | [] | 2026-03-08T03:28:51Z | 2026-04-27T04:59:52Z | 2026-04-24T20:31:00Z | CONTRIBUTOR | null | 20260427T090020Z | 2026-04-27T09:00:20Z | baccula | 22,080,883 | MDQ6VXNlcjIyMDgwODgz | User | false |
openclaw/openclaw | 4,040,324,191 | I_kwDOQb6kR87w0nRf | 39,418 | https://github.com/openclaw/openclaw/issues/39418 | https://api.github.com/repos/openclaw/openclaw/issues/39418 | Model fallback 机制在 qwen-portal token 过期时未触发 | ## 问题描述
当 qwen-portal 的 OAuth token 过期/失效时，配置的 fallback 模型未被自动调用，导致所有请求失败。
## 环境
- OpenClaw 版本: 2026.3.2
- 模型配置:
- primary: qwen-portal/coder-model
- fallbacks: [minimax-portal/MiniMax-M2.1, minimax-portal/MiniMax-M2.5, ...]
## 复现步骤
1. qwen-portal OAuth token 过期
2. 发送任何消息
3. 观察日志: 只尝试了 2 个模型 (qwen-portal/vision-m... | closed | not_planned | true | 5 | [
"stale"
] | [] | 2026-03-08T03:30:20Z | 2026-04-27T04:59:51Z | 2026-04-24T20:31:03Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | 3chet | 78,001,176 | MDQ6VXNlcjc4MDAxMTc2 | User | false |
openclaw/openclaw | 4,040,347,042 | I_kwDOQb6kR87w0s2i | 39,426 | https://github.com/openclaw/openclaw/issues/39426 | https://api.github.com/repos/openclaw/openclaw/issues/39426 | [Bug]: `openclaw agent --agent main --message "hello"` fails with ByteString error for U+2502 (│) on macOS | ### Bug type
Behavior bug (incorrect output/state without crash)
### Summary
A minimal local agent invocation fails with a ByteString conversion error on macOS.
### Minimal repro
```bash
TERM=dumb NO_COLOR=1 openclaw agent --agent main --message "hello"
### Steps to reproduce
1. Start OpenClaw on macOS.
2. Ru... | closed | completed | true | 3 | [
"bug",
"stale",
"bug:behavior"
] | [] | 2026-03-08T03:41:11Z | 2026-04-27T04:59:50Z | 2026-04-24T20:31:07Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | spaceway007 | 63,239,860 | MDQ6VXNlcjYzMjM5ODYw | User | false |
openclaw/openclaw | 4,040,354,819 | I_kwDOQb6kR87w0uwD | 39,428 | https://github.com/openclaw/openclaw/issues/39428 | https://api.github.com/repos/openclaw/openclaw/issues/39428 | Discord account-to-agent routing falls back to default agent instead of matching by account name | ## Description
When a Discord channel account name matches an agent ID, inbound messages should automatically route to that agent. Instead, they fall back to the default (main) agent, requiring explicit `bindings` config to fix.
## Steps to Reproduce
1. Configure two Discord accounts in `channels.discord.accounts`:
... | open | null | false | 2 | [] | [] | 2026-03-08T03:45:23Z | 2026-05-04T07:01:18Z | null | NONE | null | 20260504T090105Z | 2026-05-04T09:01:05Z | davidweigao | 1,740,687 | MDQ6VXNlcjE3NDA2ODc= | User | false |
openclaw/openclaw | 4,313,514,260 | I_kwDOQb6kR88AAAABARsBFA | 70,476 | https://github.com/openclaw/openclaw/issues/70476 | https://api.github.com/repos/openclaw/openclaw/issues/70476 | Android: unknown requestId error during pairing (Advance Setup, no Google Play) | ## Describe the bug
Using latest Android source code (v2026.4.21) built from source. In mainland China without Google Play services, using **Advance Setup** flow:
1. Run `openclaw qr` on MacBook to get Setup Code
2. Paste Setup Code in Android app
3. Next → connect step → approve requestId per guide
4. After approval, ... | closed | completed | true | 1 | [] | [] | 2026-04-23T04:36:10Z | 2026-04-29T04:42:05Z | 2026-04-26T07:14:12Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | mmhzlrj | 29,726,050 | MDQ6VXNlcjI5NzI2MDUw | User | false |
openclaw/openclaw | 4,313,463,370 | I_kwDOQb6kR88AAAABARo6Sg | 70,472 | https://github.com/openclaw/openclaw/issues/70472 | https://api.github.com/repos/openclaw/openclaw/issues/70472 | [Bug]: Context overflow / compaction can orphan agent:main:main and silently rotate WebChat to a new session | ## Summary
Long-running WebChat main sessions can silently lose the live `agent:main:main` mapping after context overflow / compaction-path failures, causing the next visible user message to start a new session id even though the prior transcript still exists on disk.
From the user's perspective, this looks like the ... | open | null | false | 3 | [] | [] | 2026-04-23T04:21:26Z | 2026-05-03T10:41:26Z | null | NONE | null | 20260503T120038Z | 2026-05-03T12:00:38Z | rjdjohnston | 293,686 | MDQ6VXNlcjI5MzY4Ng== | User | false |
openclaw/openclaw | 4,313,553,964 | I_kwDOQb6kR88AAAABARucLA | 70,481 | https://github.com/openclaw/openclaw/issues/70481 | https://api.github.com/repos/openclaw/openclaw/issues/70481 | [Bug]: active-memory timeoutMs not enforced — runs 30-40s despite 5s config | ## Summary
The \ctive-memory\ plugin's \ imeoutMs\ configuration is not being enforced. Despite setting \ imeoutMs: 5000\ (5 seconds), the plugin consistently runs for 30-40 seconds before timing out, blocking chat replies.
## Steps to Reproduce
1. Enable active-memory plugin with \ imeoutMs: 5000\
2. Send any mess... | closed | completed | true | 2 | [] | [] | 2026-04-23T04:47:32Z | 2026-04-28T04:42:54Z | 2026-04-26T02:26:50Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Barry-886 | 269,722,164 | U_kgDOEBOiNA | User | false |
openclaw/openclaw | 4,313,555,323 | I_kwDOQb6kR88AAAABARuhew | 70,482 | https://github.com/openclaw/openclaw/issues/70482 | https://api.github.com/repos/openclaw/openclaw/issues/70482 | [Bug]: opencode-go vision-capable models cannot use image tool — missing MediaUnderstandingProvider | ## 🐛 Bug Report (한국어 요약 아래)
### Bug type
Behavior bug (incorrect output/state without crash)
### Summary
Users subscribed to the **OpenCode Go provider** (via `opencode-go`) cannot use the `image` tool with vision-capable models (kimi-k2.6, minimax-m2.7, glm-5, mimo-v2-omni, qwen3.6-plus, etc.) even though these mod... | closed | completed | true | 1 | [] | [] | 2026-04-23T04:47:51Z | 2026-04-27T04:35:05Z | 2026-04-25T03:50:23Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | byungskers | 278,556,491 | U_kgDOEJpvSw | User | false |
openclaw/openclaw | 4,313,566,805 | I_kwDOQb6kR88AAAABARvOVQ | 70,483 | https://github.com/openclaw/openclaw/issues/70483 | https://api.github.com/repos/openclaw/openclaw/issues/70483 | Feature Request: Support excludePrefixes for session visibility filtering | ### Problem
Memory Dreaming system creates background sessions with names like:
- `dreaming-narrative-deep-xxx`
- `dreaming-narrative-rem-xxx`
These sessions clutter the session list in Control UI, making it harder to find actual user sessions.
### Current Behavior
`tools.sessions.visibility` only accepts: `self` |... | closed | completed | true | 1 | [] | [] | 2026-04-23T04:51:00Z | 2026-04-28T04:47:52Z | 2026-04-25T19:23:14Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | ChinQAQ | 152,840,343 | U_kgDOCRwolw | User | false |
openclaw/openclaw | 4,313,525,003 | I_kwDOQb6kR88AAAABARsrCw | 70,478 | https://github.com/openclaw/openclaw/issues/70478 | https://api.github.com/repos/openclaw/openclaw/issues/70478 | [Feature]: track remaining Codex harness parity and hook-policy follow-ups | ## Summary
Track the remaining Codex app-server parity and policy follow-ups after the landed native hook stack.
## Problem to solve
The main Codex parity stack is now landed:
- https://github.com/openclaw/openclaw/pull/70307
- https://github.com/openclaw/openclaw/pull/70312
- https://github.com/openclaw/openclaw/pull... | closed | duplicate | true | 0 | [
"enhancement",
"agents",
"maintainer",
"codex"
] | [
"vincentkoc"
] | 2026-04-23T04:39:03Z | 2026-04-26T04:30:30Z | 2026-04-23T04:51:29Z | MEMBER | null | 20260426T060034Z | 2026-04-26T06:00:34Z | vincentkoc | 25,068 | MDQ6VXNlcjI1MDY4 | User | false |
openclaw/openclaw | 4,213,332,756 | I_kwDOQb6kR877IlsU | 62,066 | https://github.com/openclaw/openclaw/issues/62066 | https://api.github.com/repos/openclaw/openclaw/issues/62066 | Add requireMention: "strict" mode to disable implicit thread mention bypass | ## Feature Request
### Problem
When `requireMention: true` is set for a Slack channel/account, agents correctly require an explicit `@mention` to respond in group channels. However, once an agent replies in a Slack thread, **every subsequent message in that thread bypasses the mention gate** because OpenClaw treats a... | open | null | false | 2 | [] | [] | 2026-04-06T18:24:27Z | 2026-05-02T06:31:12Z | null | NONE | null | 20260502T090030Z | 2026-05-02T09:00:30Z | yarn-rp | 61,998,848 | MDQ6VXNlcjYxOTk4ODQ4 | User | false |
openclaw/openclaw | 4,039,993,848 | I_kwDOQb6kR87wzWn4 | 39,289 | https://github.com/openclaw/openclaw/issues/39289 | https://api.github.com/repos/openclaw/openclaw/issues/39289 | [Bug]: Duplicate messages in Webchat | ### Bug type
Behavior bug (incorrect output/state without crash)
### Summary
Problem: Duplicate messages in Webchat
System:
Raspberry Pi 5, Debian 13
OpenClaw 2026.3.2
Only 1 Gateway process active
Webchat connection is stable
What happens:
My responses are displayed twice, even though only one message is sent. Ex... | closed | completed | true | 6 | [
"bug",
"bug:behavior"
] | [] | 2026-03-08T00:31:13Z | 2026-04-30T04:50:58Z | 2026-04-25T05:38:50Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | assistant-marlon | 264,953,047 | U_kgDOD8rc1w | User | false |
openclaw/openclaw | 4,313,619,411 | I_kwDOQb6kR88AAAABARyb0w | 70,487 | https://github.com/openclaw/openclaw/issues/70487 | https://api.github.com/repos/openclaw/openclaw/issues/70487 | OpenAI image generation provider doesn't support Azure OpenAI endpoints | ## Context
OpenClaw recently upgraded the default OpenAI image model to `gpt-image-2`, which is available both on `api.openai.com` and on Azure OpenAI deployments (GA'd 2026‑04‑21, doc: https://learn.microsoft.com/azure/ai-foundry/openai/how-to/dall-e ). A lot of enterprise users route OpenAI traffic through an Azure ... | closed | completed | true | 0 | [] | [] | 2026-04-23T05:03:45Z | 2026-04-26T04:30:22Z | 2026-04-23T11:36:58Z | CONTRIBUTOR | null | 20260426T060034Z | 2026-04-26T06:00:34Z | ottodeng | 20,509,364 | MDQ6VXNlcjIwNTA5MzY0 | User | false |
openclaw/openclaw | 4,313,514,062 | I_kwDOQb6kR88AAAABARsATg | 70,475 | https://github.com/openclaw/openclaw/issues/70475 | https://api.github.com/repos/openclaw/openclaw/issues/70475 | Feature: Configure dedicated LLM model for Dreaming subsystem | ## Summary
Request the ability to configure a separate/dedicated LLM model for the Dreaming subsystem (REM / Light / Deep phases), instead of inheriting the agent's primary model.
## Problem
Currently, Dreaming uses the same model as the agent (e.g. `ollama/glm-5.1:cloud`). However, Dreaming tasks — claim extraction, ... | closed | completed | true | 2 | [] | [] | 2026-04-23T04:36:06Z | 2026-05-01T04:46:05Z | 2026-04-28T05:02:36Z | NONE | null | 20260501T113057Z | 2026-05-01T11:30:57Z | c0d1ngHUB | 268,152,820 | U_kgDOD_uv9A | User | false |
openclaw/openclaw | 4,313,589,385 | I_kwDOQb6kR88AAAABARwmiQ | 70,485 | https://github.com/openclaw/openclaw/issues/70485 | https://api.github.com/repos/openclaw/openclaw/issues/70485 | [Bug]: 2026.4.22 passes tools: [] to Pi embedded sessions and disables customTools | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
1. Start from a pre-2026.4.22 OpenClaw build where Pi embedded tool calling works.
2. Update to OpenClaw 2026.4.22.
3. Run turns that require local/custom tool execution through Pi embedded sessions.
4. Observe that usable t... | closed | completed | true | 2 | [
"bug",
"regression"
] | [] | 2026-04-23T04:56:33Z | 2026-04-26T04:30:16Z | 2026-04-23T16:36:33Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | colmbrogan | 73,212,305 | MDQ6VXNlcjczMjEyMzA1 | User | false |
openclaw/openclaw | 4,039,358,218 | I_kwDOQb6kR87ww7cK | 39,156 | https://github.com/openclaw/openclaw/issues/39156 | https://api.github.com/repos/openclaw/openclaw/issues/39156 | [Bug]: diagnostics-otel plugin: OTLP log export doesn't work — module isolation between gateway and plugin-sdk | ### Bug type
Behavior bug (incorrect output/state without crash)
### Summary
Summary
The diagnostics-otel plugin's log export feature (diagnostics.otel.logs: true) doesn't work. The plugin loads successfully and reports "logs exporter enabled (OTLP/Protobuf)" but no log records are ever sent to the OTLP endpoint. M... | closed | completed | true | 7 | [
"bug",
"bug:behavior"
] | [] | 2026-03-07T20:07:25Z | 2026-04-30T04:50:59Z | 2026-04-25T05:38:37Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | gilmichlin | 838,384 | MDQ6VXNlcjgzODM4NA== | User | false |
openclaw/openclaw | 4,313,269,704 | I_kwDOQb6kR88AAAABARdFyA | 70,457 | https://github.com/openclaw/openclaw/issues/70457 | https://api.github.com/repos/openclaw/openclaw/issues/70457 | [Bug]: npm install -g openclaw@latest fails with missing module '@larksuiteoapi/node-sdk' and 'nostr-tools' on fresh macOS and Ubuntu environments | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Fresh installation using `npm install -g openclaw@latest` fails during initial setup because required extension dependencies such as `@larksuiteoapi/node-sdk` and `nostr-tools` are missing from the global package installatio... | closed | completed | true | 26 | [
"bug",
"regression"
] | [] | 2026-04-23T03:24:35Z | 2026-04-26T04:30:13Z | 2026-04-23T16:44:03Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | zhaogewudi666 | 85,532,106 | MDQ6VXNlcjg1NTMyMTA2 | User | false |
openclaw/openclaw | 4,313,683,232 | I_kwDOQb6kR88AAAABAR2VIA | 70,489 | https://github.com/openclaw/openclaw/issues/70489 | https://api.github.com/repos/openclaw/openclaw/issues/70489 | Feature Request: Allow memory-core to run alongside external memory plugins for dreaming | # Feature Request: Allow memory-core to run alongside external memory plugins (e.g., memory-lancedb-pro) for dreaming functionality
## Problem Description
Currently, OpenClaw's memory slot mechanism uses an **exclusive allocation** strategy, meaning only one plugin can occupy the `memory` slot at a time. This prevent... | closed | completed | true | 2 | [] | [] | 2026-04-23T05:18:17Z | 2026-04-28T04:47:51Z | 2026-04-25T19:23:20Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | LittleRollo | 256,484,611 | U_kgDOD0mlAw | User | false |
openclaw/openclaw | 4,313,820,088 | I_kwDOQb6kR88AAAABAR-ruA | 70,492 | https://github.com/openclaw/openclaw/issues/70492 | https://api.github.com/repos/openclaw/openclaw/issues/70492 | Bug: openclaw status / status --json hang or die under 2026.4.21 due to heavy status runtime/plugin load path | # Bug: `openclaw status` / `openclaw status --json` hang or die under 2026.4.21 due to heavy status runtime/plugin load path
## Summary
On OpenClaw `2026.4.21 (f788c88)`, `openclaw status` and `openclaw status --json` are unreliable in this environment.
Symptoms:
- command hangs for a long time
- child `openclaw` pro... | closed | completed | true | 1 | [] | [] | 2026-04-23T05:50:17Z | 2026-04-28T04:47:50Z | 2026-04-25T19:23:27Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | iaskynet0-code | 271,433,290 | U_kgDOEC2-Sg | User | false |
openclaw/openclaw | 4,313,825,305 | I_kwDOQb6kR88AAAABAR_AGQ | 70,493 | https://github.com/openclaw/openclaw/issues/70493 | https://api.github.com/repos/openclaw/openclaw/issues/70493 | Isolated session gateway missing full Playwright support | ## Problem
When using the browser tool in an isolated sub-agent session, Playwright operations fail with:
error: Playwright is not available in this gateway build; 'navigate' is unsupported.
Install the full Playwright package (not playwright-core) and restart the gateway.
Main session works fine with profile=user +... | open | null | false | 3 | [] | [] | 2026-04-23T05:51:28Z | 2026-05-03T07:45:02Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | xdengli | 149,284,229 | U_kgDOCOXlhQ | User | false |
openclaw/openclaw | 4,020,072,527 | I_kwDOQb6kR87vnXBP | 33,960 | https://github.com/openclaw/openclaw/issues/33960 | https://api.github.com/repos/openclaw/openclaw/issues/33960 | [Feature]: Image generation input parity for URL and path sources | ## Summary
Request a consistent image-input contract for image generation tools that supports both remote URLs and local paths, with clear behavior per channel/runtime.
## Why
Some deployments need URL-based image access (for hosted assets) instead of path-only flows. Current expectations are inconsistent across tool ... | closed | completed | true | 0 | [
"maintainer"
] | [
"vincentkoc"
] | 2026-03-04T05:08:31Z | 2026-04-26T04:30:29Z | 2026-04-23T05:52:20Z | MEMBER | null | 20260426T060034Z | 2026-04-26T06:00:34Z | vincentkoc | 25,068 | MDQ6VXNlcjI1MDY4 | User | false |
openclaw/openclaw | 4,010,704,618 | I_kwDOQb6kR87vDn7q | 31,618 | https://github.com/openclaw/openclaw/issues/31618 | https://api.github.com/repos/openclaw/openclaw/issues/31618 | [Feature]: support sending images directly instead of URLs | ### Summary
When using the Feishu (Lark) channel, the bot currently sends image URLs as plain text links instead of actual images.
Expected behavior: The bot should download the image and upload it to Feishu using the im:resource API, then send it as a proper image message using image_key.
We already have im:resourc... | closed | completed | true | 1 | [
"enhancement"
] | [
"vincentkoc"
] | 2026-03-02T11:17:44Z | 2026-04-26T04:30:29Z | 2026-04-23T05:52:34Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | vaeying151-design | 247,341,776 | U_kgDODr4i0A | User | false |
openclaw/openclaw | 4,313,818,147 | I_kwDOQb6kR88AAAABAR-kIw | 70,491 | https://github.com/openclaw/openclaw/issues/70491 | https://api.github.com/repos/openclaw/openclaw/issues/70491 | Control UI status bar shows '0% ctx' when actual session context is 44% — contextPercent ignores cacheRead | # Bug: Control UI status bar `% ctx` shows 0% even when server-side context is 44%
**Reporter:** Giang Tran
**Date:** 2026-04-23
**OpenClaw version:** `2026.4.21` (build `f788c88`)
**Channel:** Webchat (Control UI)
---
## Summary
The Control UI message-status bar shows `0% ctx` even when the actual session context ... | closed | completed | true | 3 | [] | [] | 2026-04-23T05:49:57Z | 2026-04-26T04:30:06Z | 2026-04-23T17:13:33Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | giangthb | 45,944,331 | MDQ6VXNlcjQ1OTQ0MzMx | User | false |
openclaw/openclaw | 4,313,859,180 | I_kwDOQb6kR88AAAABASBEbA | 70,502 | https://github.com/openclaw/openclaw/issues/70502 | https://api.github.com/repos/openclaw/openclaw/issues/70502 | Gateway enters infinite error loop: TypeError: Cannot read properties of undefined (reading 'startsWith') | ## Bug Report
**Version:** 2026.4.12 (1c0672b)
**Severity:** High — gateway becomes completely unresponsive, consuming 100%+ CPU
### Error
```
TypeError: Cannot read properties of undefined (reading 'startsWith')
```
**Location:** `dist/subsystem-CSBs8ZPt.js:161` (minified)
**Triggered by:** Gateway request handler... | closed | completed | true | 0 | [] | [] | 2026-04-23T05:59:35Z | 2026-04-26T04:30:04Z | 2026-04-23T18:51:56Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | DoobyTheHouseElf | 278,503,277 | U_kgDOEJmfbQ | User | false |
openclaw/openclaw | 4,313,897,066 | I_kwDOQb6kR88AAAABASDYag | 70,505 | https://github.com/openclaw/openclaw/issues/70505 | https://api.github.com/repos/openclaw/openclaw/issues/70505 | [Bug]: Control UI: assistant-media preview returns 401 in device-token sessions even when chat works | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
In Control UI, assistant-generated media attachments (for example `MEDIA:/...png`) fail to render with `401 Unauthorized`, even though the chat session itself is working.
This appears to be an auth mismatch be... | closed | completed | true | 2 | [
"bug",
"bug:behavior"
] | [] | 2026-04-23T06:08:26Z | 2026-04-26T04:29:41Z | 2026-04-23T23:31:12Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | pavelknezik | 274,213,128 | U_kgDOEFgpCA | User | false |
openclaw/openclaw | 4,205,624,592 | I_kwDOQb6kR876rL0Q | 61,011 | https://github.com/openclaw/openclaw/issues/61011 | https://api.github.com/repos/openclaw/openclaw/issues/61011 | [Bug] Tool execution fails silently — I claim to run commands but nothing happens | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
When I say I'll execute a command (transcribe audio, write files, run scripts), I respond with acknowledgment ("Let me do that..." / "Writing it now...") but the tool never actually executes. No output, no error, no result
... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-04T19:11:13Z | 2026-04-26T04:30:28Z | 2026-04-23T06:15:10Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | dylanpassey-hub | 263,126,731 | U_kgDOD67-yw | User | false |
openclaw/openclaw | 4,313,921,234 | I_kwDOQb6kR88AAAABASE20g | 70,507 | https://github.com/openclaw/openclaw/issues/70507 | https://api.github.com/repos/openclaw/openclaw/issues/70507 | [Bug]: Control UI: user-uploaded chat images disappear from history after send | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
## Summary
When a user inserts an image into Control UI chat, the image is visible immediately after sending, but disappears afterward and is no longer shown in chat history.
## Symptoms
- User pastes/drops an image into chat
- UI sho... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-23T06:13:55Z | 2026-04-28T04:47:49Z | 2026-04-25T19:23:34Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | pavelknezik | 274,213,128 | U_kgDOEFgpCA | User | false |
openclaw/openclaw | 4,313,956,411 | I_kwDOQb6kR88AAAABASHAOw | 70,509 | https://github.com/openclaw/openclaw/issues/70509 | https://api.github.com/repos/openclaw/openclaw/issues/70509 | [Bug] Dreaming raw candidates pollute daily memory files (memory/YYYY-MM-DD.md) | ## Problem Summary
The dreaming subsystem's fallback mechanism (triggered when subagent runtime is unavailable in cron/system-event context) dumps raw memory candidates into daily files alongside curated human-readable content.
This appears related to #68723 (which fixed the same issue for ), but the daily memory fi... | closed | completed | true | 1 | [] | [] | 2026-04-23T06:21:37Z | 2026-04-28T04:47:48Z | 2026-04-25T19:23:41Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | dylanpassey-hub | 263,126,731 | U_kgDOD67-yw | User | false |
openclaw/openclaw | 4,104,671,791 | I_kwDOQb6kR870qFIv | 50,686 | https://github.com/openclaw/openclaw/issues/50686 | https://api.github.com/repos/openclaw/openclaw/issues/50686 | [Feature]: Zoom Team Chat Channel Integration | ### Summary
Summary
Adds full support for Zoom Team Chat as a core channel plugin, enabling Moltbot to receive and respond to direct messages via Zoom's Team Chat Bot API.
Closed PR here: https://github.com/openclaw/openclaw/pull/3279
Contributor Background: Implemented by [@pranavk95](https://www.linkedin.com/in/pr... | closed | not_planned | true | 4 | [
"enhancement"
] | [] | 2026-03-19T22:57:01Z | 2026-04-28T04:54:41Z | 2026-04-25T04:53:03Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | pranavk95 | 11,080,417 | MDQ6VXNlcjExMDgwNDE3 | User | false |
openclaw/openclaw | 4,313,972,435 | I_kwDOQb6kR88AAAABASH-0w | 70,512 | https://github.com/openclaw/openclaw/issues/70512 | https://api.github.com/repos/openclaw/openclaw/issues/70512 | Feature Request: Feishu DM thread/topic-based session isolation | ## Feature Request
**Channel:** Feishu (Lark)
**Current behavior:** All DM messages from the same user route to a single session, regardless of Feishu topic/thread ID.
**Requested behavior:** Support per-topic session isolation in Feishu DMs, similar to how group topics already get separate sessions.
### Problem
Whe... | open | null | false | 1 | [] | [] | 2026-04-23T06:25:49Z | 2026-05-02T22:24:26Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | cjboy007 | 64,246,343 | MDQ6VXNlcjY0MjQ2MzQz | User | false |
openclaw/openclaw | 4,313,969,916 | I_kwDOQb6kR88AAAABASH0_A | 70,511 | https://github.com/openclaw/openclaw/issues/70511 | https://api.github.com/repos/openclaw/openclaw/issues/70511 | [Bug]: native codex harness uses broken isolated CODEX_HOME (harness-auth) while ~/.codex works | ## Summary
When using the native Codex harness (`codex/gpt-5.4` with `embeddedHarness.runtime: "codex"`), OpenClaw launches Codex app-server with an isolated `CODEX_HOME` under:
`~/.openclaw/agents/main/agent/harness-auth/codex/<hash>`
On my machine, that generated home is the failing layer.
- Real Codex auth under... | closed | completed | true | 2 | [] | [] | 2026-04-23T06:25:09Z | 2026-04-28T04:47:47Z | 2026-04-25T19:23:47Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | amart-builder | 264,521,053 | U_kgDOD8RFXQ | User | false |
openclaw/openclaw | 4,314,007,670 | I_kwDOQb6kR88AAAABASKIdg | 70,516 | https://github.com/openclaw/openclaw/issues/70516 | https://api.github.com/repos/openclaw/openclaw/issues/70516 | [Bug]: WhatsApp channel login fails | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
Yes
### Summary
### **Channel reports:**
****Channel whatsapp does not support login**
Onboarding cannot continue**
Environment
OS: Ubuntu (WSL)
Node.js: v24.13.1
npm: 11.8.0
OpenClaw: 2026.4.21 (f788c88)
_**missing @openclaw/whatsapp pac... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-23T06:33:34Z | 2026-04-27T05:09:40Z | 2026-04-24T05:42:01Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | FattyBull | 16,086,854 | MDQ6VXNlcjE2MDg2ODU0 | User | false |
openclaw/openclaw | 4,313,962,214 | I_kwDOQb6kR88AAAABASHW5g | 70,510 | https://github.com/openclaw/openclaw/issues/70510 | https://api.github.com/repos/openclaw/openclaw/issues/70510 | [maintainer created] Thai translation feedback | Thai docs are now generated from the openclaw/docs publish repo via the GPT-5.4 docs-i18n workflow.\n\nUse this issue to report bad Thai translations, systematic terminology issues, awkward phrasing, or places where the translation prompt/glossary should be improved.\n\nPlease include:\n- the docs URL or file path\n- t... | closed | completed | true | 1 | [
"maintainer"
] | [] | 2026-04-23T06:23:08Z | 2026-04-26T04:30:25Z | 2026-04-23T06:33:44Z | CONTRIBUTOR | null | 20260426T060034Z | 2026-04-26T06:00:34Z | steipete | 58,493 | MDQ6VXNlcjU4NDkz | User | false |
openclaw/openclaw | 4,314,022,648 | I_kwDOQb6kR88AAAABASLC-A | 70,517 | https://github.com/openclaw/openclaw/issues/70517 | https://api.github.com/repos/openclaw/openclaw/issues/70517 | [Bug]: Telegram voice/audio messages have significant delivery delay (~2 minutes) before agent receives them | ### Bug type
Performance / Latency
### Beta release blocker
No
### Summary
When a Telegram user sends a voice message (OGG/Opus), the agent receives the message and processes the audio significantly later than when it was sent — observed delay of ~2 minutes. Text messages on the same channel are delivered and proc... | closed | completed | true | 1 | [] | [] | 2026-04-23T06:36:29Z | 2026-04-28T04:47:45Z | 2026-04-25T19:24:00Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | LuisAmaury | 33,032,099 | MDQ6VXNlcjMzMDMyMDk5 | User | false |
openclaw/openclaw | 4,159,255,794 | I_kwDOQb6kR8736TTy | 56,344 | https://github.com/openclaw/openclaw/issues/56344 | https://api.github.com/repos/openclaw/openclaw/issues/56344 | [Feature]: Enable generic file upload support in openclaw-control-ui | ### Summary
Currently, the file upload component in the openclaw-control-ui interface is restricted to image formats (MIME type image/*). This limitation restricts the efficiency of users who need to transmit documents, code snippets, or logs directly through the web console.
### Problem to solve
Allowing users to u... | closed | not_planned | true | 3 | [
"enhancement"
] | [] | 2026-03-28T10:39:55Z | 2026-04-29T04:40:47Z | 2026-04-26T08:43:08Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | wzq0824 | 81,143,005 | MDQ6VXNlcjgxMTQzMDA1 | User | false |
openclaw/openclaw | 4,173,804,364 | I_kwDOQb6kR874xzNM | 57,920 | https://github.com/openclaw/openclaw/issues/57920 | https://api.github.com/repos/openclaw/openclaw/issues/57920 | Stale ended subagent runs persist in childSessions, ghost agents on Floor | ## Bug
Ended subagent runs accumulate in `~/.openclaw/subagents/runs.json` indefinitely and appear as ghost child sessions on the Floor UI.
## Root Cause
`resolveChildSessionKeys()` (in `auth-profiles-*.js`, around the subagent-registry region) iterates all runs for a controller session key but **never filters by `e... | closed | completed | true | 4 | [] | [] | 2026-03-30T20:16:33Z | 2026-04-30T04:55:33Z | 2026-04-25T04:39:24Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | piergiuliol | 11,920,903 | MDQ6VXNlcjExOTIwOTAz | User | false |
openclaw/openclaw | 4,314,202,711 | I_kwDOQb6kR88AAAABASWCVw | 70,520 | https://github.com/openclaw/openclaw/issues/70520 | https://api.github.com/repos/openclaw/openclaw/issues/70520 | Gateway image: support mounting $HOME on a persistent volume (seed image defaults on first boot) | ## Problem
When OpenClaw gateway is deployed to a platform that gives the container a single persistent volume (Fly.io, Render, fly-style K8s persistent volumes, etc.), everything outside that volume is ephemeral. But meaningful agent state lives under `$HOME`:
- Default agent workspace — `$HOME/.openclaw/workspace/`... | closed | completed | true | 1 | [] | [] | 2026-04-23T07:13:01Z | 2026-04-26T04:30:25Z | 2026-04-23T07:13:43Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | longfin | 128,436 | MDQ6VXNlcjEyODQzNg== | User | false |
openclaw/openclaw | 4,314,209,357 | I_kwDOQb6kR88AAAABASWcTQ | 70,521 | https://github.com/openclaw/openclaw/issues/70521 | https://api.github.com/repos/openclaw/openclaw/issues/70521 | Chicken-and-egg: `openclaw doctor --fix` cannot repair missing staged runtime deps because config loading itself fails | ## Summary
After upgrading OpenClaw (e.g. 2026.4.12 → 2026.4.21), bundled plugin staged runtime dependencies (such as `@larksuiteoapi/node-sdk` for the feishu plugin) may be missing. Running `openclaw doctor --fix` should repair this, but instead fails with `Cannot find module '@larksuiteoapi/node-sdk'` during config ... | closed | completed | true | 1 | [] | [] | 2026-04-23T07:14:21Z | 2026-04-26T04:30:10Z | 2026-04-23T16:47:10Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | ggzeng | 20,488,795 | MDQ6VXNlcjIwNDg4Nzk1 | User | false |
openclaw/openclaw | 4,314,235,604 | I_kwDOQb6kR88AAAABASYC1A | 70,524 | https://github.com/openclaw/openclaw/issues/70524 | https://api.github.com/repos/openclaw/openclaw/issues/70524 | [Bug] Discord: adding new channel to allowlist requires gateway restart (hot-reload ineffective) | ## Description
When adding a new Discord channel ID to the allowlist configuration, hot-reload does not pick up the change. The gateway must be fully restarted for the new channel to become active.
## Steps to Reproduce
1. Create a new Discord channel in the server
2. Add the channel ID to Discord allowlist in confi... | closed | completed | true | 1 | [] | [] | 2026-04-23T07:19:22Z | 2026-04-28T04:47:45Z | 2026-04-25T19:24:07Z | CONTRIBUTOR | null | 20260428T060020Z | 2026-04-28T06:00:20Z | kagura-agent | 268,167,063 | U_kgDOD_vnlw | User | false |
openclaw/openclaw | 4,314,297,048 | I_kwDOQb6kR88AAAABASby2A | 70,526 | https://github.com/openclaw/openclaw/issues/70526 | https://api.github.com/repos/openclaw/openclaw/issues/70526 | [Regression][2026.4.21] npm global install still triggers bundled extension MODULE_NOT_FOUND | ### Bug type
Regression (worked before, now fails)
### Summary
On OpenClaw `2026.4.21`, a fresh global npm install still causes cascading `MODULE_NOT_FOUND` errors from bundled extensions, and `openclaw status` / `openclaw health` become unreliable.
### Environment
- OpenClaw: `2026.4.21`
- Install method: `npm insta... | closed | completed | true | 2 | [] | [] | 2026-04-23T07:31:45Z | 2026-04-26T04:30:12Z | 2026-04-23T16:44:08Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | liaoandi | 31,417,269 | MDQ6VXNlcjMxNDE3MjY5 | User | false |
openclaw/openclaw | 4,314,334,024 | I_kwDOQb6kR88AAAABASeDSA | 70,529 | https://github.com/openclaw/openclaw/issues/70529 | https://api.github.com/repos/openclaw/openclaw/issues/70529 | [Bug]: Desktop cannot use existing Chrome sessions: EasyClaw Google sign-in fails, and user profile attach fails with spawn npx ENOENT | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
On Genspark Claw Desktop, two paths to use existing Chrome sessions both fail: (1) EasyClaw Browser Relay extension Google sign-in returns an error after OAuth callback, (2) built-in `user` profile attach fails... | open | null | false | 2 | [
"bug"
] | [] | 2026-04-23T07:38:46Z | 2026-05-02T12:37:21Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | yuzo-machiya | 278,620,287 | U_kgDOEJtofw | User | false |
openclaw/openclaw | 4,314,333,953 | I_kwDOQb6kR88AAAABASeDAQ | 70,528 | https://github.com/openclaw/openclaw/issues/70528 | https://api.github.com/repos/openclaw/openclaw/issues/70528 | [Feature]: Automatic rollback when agent-generated config changes break the system |
## Summary
In autonomous workflows, OpenClaw agents may modify config/env values automatically.
When those generated changes are invalid, the system can become unstable or fail to start, requiring manual recovery.
## Problem
This is more serious than normal user config mistakes, because the failure is created dur... | closed | completed | true | 1 | [
"enhancement"
] | [] | 2026-04-23T07:38:45Z | 2026-04-28T04:47:44Z | 2026-04-25T19:24:14Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | johnsonbuilds | 48,879,766 | MDQ6VXNlcjQ4ODc5NzY2 | User | false |
openclaw/openclaw | 4,314,302,136 | I_kwDOQb6kR88AAAABAScGuA | 70,527 | https://github.com/openclaw/openclaw/issues/70527 | https://api.github.com/repos/openclaw/openclaw/issues/70527 | memory_search can fail with `Cannot read properties of undefined (reading 'get')` due to corrupted shared singleton cache | # memory_search can fail with `Cannot read properties of undefined (reading 'get')` due to corrupted shared singleton cache
## Summary
`memory_search` can become unavailable in the long-running gateway process even while fresh CLI invocations still work.
The user-visible tool result is:
```json
{
"results": [],
... | closed | completed | true | 1 | [] | [] | 2026-04-23T07:32:37Z | 2026-04-26T04:29:05Z | 2026-04-24T04:21:05Z | CONTRIBUTOR | null | 20260426T060034Z | 2026-04-26T06:00:34Z | fire-mirror | 88,885,494 | MDQ6VXNlcjg4ODg1NDk0 | User | false |
openclaw/openclaw | 4,314,375,375 | I_kwDOQb6kR88AAAABASgkzw | 70,531 | https://github.com/openclaw/openclaw/issues/70531 | https://api.github.com/repos/openclaw/openclaw/issues/70531 | [Feature]: Split jobs.json into per-job config files | ### Summary
Allow cron jobs to be stored as individual JSON files (e.g., ~/.openclaw/cron/jobs/<job-id>.json) instead of a single monolithic jobs.json file.
### Problem to solve
Currently all cron jobs are stored in a single ~/.openclaw/cron/jobs.json file (~22KB with 10+ jobs). This creates several issues:
1. **R... | closed | not_planned | true | 1 | [
"enhancement"
] | [] | 2026-04-23T07:46:40Z | 2026-04-29T04:34:33Z | 2026-04-26T18:57:44Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | TheoDoreW | 26,006,532 | MDQ6VXNlcjI2MDA2NTMy | User | false |
openclaw/openclaw | 4,314,413,517 | I_kwDOQb6kR88AAAABASi5zQ | 70,533 | https://github.com/openclaw/openclaw/issues/70533 | https://api.github.com/repos/openclaw/openclaw/issues/70533 | Plugin discovery loads all dist/extensions/ manifests at boot regardless of tools.allow (~500 MB structural heap) | ## Summary
Plugin discovery reads every manifest under `dist/extensions/` via `readFileSync` and runs each through zod validation at boot — even when `tools.allow` restricts the enabled plugin set to a small subset. For an install with 104 bundled extensions and a 6-plugin allowlist, this contributes ~500 MB of reside... | open | null | false | 2 | [] | [] | 2026-04-23T07:53:49Z | 2026-05-02T13:57:04Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | jpippo364 | 256,246,144 | U_kgDOD0YBgA | User | false |
openclaw/openclaw | 4,314,424,985 | I_kwDOQb6kR88AAAABASjmmQ | 70,534 | https://github.com/openclaw/openclaw/issues/70534 | https://api.github.com/repos/openclaw/openclaw/issues/70534 | [Bug]: Dreaming fallback triggers constantly when using named agent (agents/<name>/sessions/ path not recognized) | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
When using a named agent (`--agent main` or equivalent), Dreaming's corpus
collector cannot find session data and falls back to daily memory files
permanently. This contradicts the v2026.4.10 changelog which states fal... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-23T07:55:51Z | 2026-04-28T04:47:43Z | 2026-04-25T19:24:20Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | HuangMou0828 | 46,060,300 | MDQ6VXNlcjQ2MDYwMzAw | User | false |
openclaw/openclaw | 4,313,981,467 | I_kwDOQb6kR88AAAABASIiGw | 70,513 | https://github.com/openclaw/openclaw/issues/70513 | https://api.github.com/repos/openclaw/openclaw/issues/70513 | [Feature request]: Advanced Persistent Memory & Knowledge Management System for Autonomous Agents | ### Summary
Introduce a structured and evolving memory system so agents can actually remember, organize and reuse information across sessions instead of starting from scratch every time.
### Problem to solve
Right now the system behaves like it has almost no real memory continuity. As soon as a session ends, most of... | closed | completed | true | 3 | [
"enhancement"
] | [] | 2026-04-23T06:28:03Z | 2026-04-28T04:47:46Z | 2026-04-25T19:23:54Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | alexandre-leng | 72,550,137 | MDQ6VXNlcjcyNTUwMTM3 | User | false |
openclaw/openclaw | 4,092,937,318 | I_kwDOQb6kR87z9URm | 49,515 | https://github.com/openclaw/openclaw/issues/49515 | https://api.github.com/repos/openclaw/openclaw/issues/49515 | Session idle/daily reset never fires: updateLastRoute() bumps updatedAt before freshness check | ## Bug Description
Session idle and daily resets never fire because `updateLastRoute()` in `recordInboundSession()` bumps `updatedAt` to `Date.now()` **before** `initSessionState()` evaluates session freshness.
## Environment
- OpenClaw version: 2026.3.13 (61d171a)
- Platform: macOS (Darwin arm64), Node v25.6.1
- Cha... | closed | completed | true | 3 | [] | [] | 2026-03-18T05:56:58Z | 2026-04-26T04:30:23Z | 2026-04-23T08:05:07Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | slimvaio | 4,862,654 | MDQ6VXNlcjQ4NjI2NTQ= | User | false |
openclaw/openclaw | 4,314,484,658 | I_kwDOQb6kR88AAAABASnPsg | 70,538 | https://github.com/openclaw/openclaw/issues/70538 | https://api.github.com/repos/openclaw/openclaw/issues/70538 | [Bug]: [BUG] Inconsistent/Recurring TypeError: Cannot read properties of undefined (reading 'trim') during openclaw config / plugin loading. | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Ik maak het graag nog een keer voor je, en deze keer in een professionele structuur die perfect is voor een GitHub Issue (of Jira/Bug Tracker). Ik heb alle technische details uit onze conversatie meegenomen om ... | closed | completed | true | 2 | [
"bug",
"close:already-fixed",
"dedupe:child",
"bug:behavior"
] | [
"vincentkoc"
] | 2026-04-23T08:06:07Z | 2026-04-28T05:29:08Z | 2026-04-25T09:37:04Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | coenheer-tech | 278,626,073 | U_kgDOEJt_GQ | User | false |
openclaw/openclaw | 4,314,524,975 | I_kwDOQb6kR88AAAABASptLw | 70,541 | https://github.com/openclaw/openclaw/issues/70541 | https://api.github.com/repos/openclaw/openclaw/issues/70541 | [Bug]: post-compaction context injection uses process.cwd() instead of run.workspaceDir (agent-runner.runtime line 3312) | ## Summary
`readPostCompactionContext` is called with `process.cwd()` instead of the run's configured `workspaceDir` inside the auto-compaction post-hook for follow-up runs (`agent-runner.runtime-CH0aH7T6.js` line 3312). For agents that run via a CLI provider subprocess (e.g. `openai-codex`), `process.cwd()` inside th... | open | null | false | 1 | [] | [] | 2026-04-23T08:13:15Z | 2026-05-02T12:13:13Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | benikigai | 5,900,155 | MDQ6VXNlcjU5MDAxNTU= | User | false |
openclaw/openclaw | 4,314,444,877 | I_kwDOQb6kR88AAAABASk0TQ | 70,537 | https://github.com/openclaw/openclaw/issues/70537 | https://api.github.com/repos/openclaw/openclaw/issues/70537 | [Bug]: Gateway serialization produces invalid JSON on tool calls with array fields (intermittent) | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
The gateway intermittently produces malformed JSON when serializing tool calls that contain array-typed parameters (e.g. `fields: [{...}, {...}]`). The resulting JSON fails to parse with errors like `Expected '... | closed | completed | true | 2 | [
"bug",
"bug:behavior"
] | [] | 2026-04-23T07:59:10Z | 2026-04-29T04:35:47Z | 2026-04-26T14:56:23Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | korentins | 6,073,077 | MDQ6VXNlcjYwNzMwNzc= | User | false |
openclaw/openclaw | 4,314,617,731 | I_kwDOQb6kR88AAAABASvXgw | 70,545 | https://github.com/openclaw/openclaw/issues/70545 | https://api.github.com/repos/openclaw/openclaw/issues/70545 | v2026.4.21: WhatsApp plugin fails to load — @whiskeysockets/baileys missing from distribution; watchdog reinstalls broken version without health-gate | ## Summary
Two related bugs that caused a complete WhatsApp outage on upgrade to v2026.4.21.
## Bug 1 — Baileys missing from distribution
**Version:** 2026.4.21
**Node:** v24.13.1 (plist) / v25.6.1 (running)
**Platform:** macOS (Apple Silicon)
After upgrading to v2026.4.21, the WhatsApp plugin fails to load on ... | closed | completed | true | 1 | [] | [] | 2026-04-23T08:29:08Z | 2026-04-26T04:30:08Z | 2026-04-23T16:47:17Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | azeemazhar2 | 151,832,497 | U_kgDOCQzHsQ | User | false |
openclaw/openclaw | 4,314,630,453 | I_kwDOQb6kR88AAAABASwJNQ | 70,546 | https://github.com/openclaw/openclaw/issues/70546 | https://api.github.com/repos/openclaw/openclaw/issues/70546 | loopDetection: generic_repeat detector only emits warning, never blocks execution | # loopDetection: generic_repeat detector only emits warning, never blocks execution
## Bug Description
The `loopDetection.generic_repeat` detector in `tool-loop-detection.ts` only ever returns `level: "warning"` — it never escalates to `level: "critical"`, meaning it cannot block runaway tool calls.
## Evidence
Tod... | closed | not_planned | true | 1 | [] | [] | 2026-04-23T08:31:38Z | 2026-04-29T04:42:03Z | 2026-04-26T07:16:17Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | Joni628 | 272,840,004 | U_kgDOEEM1RA | User | false |
openclaw/openclaw | 4,314,698,879 | I_kwDOQb6kR88AAAABAS0Ufw | 70,552 | https://github.com/openclaw/openclaw/issues/70552 | https://api.github.com/repos/openclaw/openclaw/issues/70552 | Heartbeat exec_completion event spawns phantom user-facing reply when command output is not attached | ## Summary
The heartbeat `exec_completion` handler can spawn a user-facing agent turn with no command output attached, producing phantom replies like *"I don't see any system messages above with command output to relay..."* delivered to end users on Telegram/Discord/etc.
## Environment
- OpenClaw `2026.4.21`
- Chann... | closed | completed | true | 1 | [] | [] | 2026-04-23T08:45:13Z | 2026-04-28T04:47:42Z | 2026-04-25T19:24:27Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | danielt69 | 3,322,065 | MDQ6VXNlcjMzMjIwNjU= | User | false |
openclaw/openclaw | 4,314,796,769 | I_kwDOQb6kR88AAAABAS6S4Q | 70,553 | https://github.com/openclaw/openclaw/issues/70553 | https://api.github.com/repos/openclaw/openclaw/issues/70553 | fix(discord): Cannot access rawData on partial Channel when executing slash commands in thread channels | # Bug: `Cannot access rawData on partial Channel` when executing slash commands in thread channels
## Environment
- OpenClaw version: `2026.4.21`
- Carbon version: `latest` (bundled in `dist/extensions/discord/node_modules/@buape/carbon`)
- Node.js: `v24.13.0`
- OS: Windows (likely affects all platforms)
## Error Me... | closed | completed | true | 1 | [] | [] | 2026-04-23T09:02:28Z | 2026-04-26T04:30:16Z | 2026-04-23T16:36:33Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | jlin53882 | 76,562,059 | MDQ6VXNlcjc2NTYyMDU5 | User | false |
openclaw/openclaw | 4,314,969,254 | I_kwDOQb6kR88AAAABATE0pg | 70,558 | https://github.com/openclaw/openclaw/issues/70558 | https://api.github.com/repos/openclaw/openclaw/issues/70558 | [Bug]: Bug: User message appears twice in webchat (session JSONL correct, UI renders duplicate) | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
## Bug: User message appears twice in webchat
### Environment
- OpenClaw: 2026.4.20 / 2026.4.21
- Channel: webchat (Control UI in browser, single tab)
- OS: Windows
### Steps to reproduce
1. Send any message via Control ... | closed | completed | true | 1 | [
"bug"
] | [] | 2026-04-23T09:33:01Z | 2026-04-28T04:47:39Z | 2026-04-25T19:24:54Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | heixingyuanfang | 273,284,829 | U_kgDOEEn-3Q | User | false |
openclaw/openclaw | 4,314,967,306 | I_kwDOQb6kR88AAAABATEtCg | 70,556 | https://github.com/openclaw/openclaw/issues/70556 | https://api.github.com/repos/openclaw/openclaw/issues/70556 | [Bug]: webchat I can not stop openclaw? | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
[Bug]: webchat I can not stop openclaw? when send /stop not work,please help me
### Steps to reproduce
talk in webchat many times ,after 10 min ,send /stop it does not work,stop failed.
### Expected behavior
stop succ
#... | closed | completed | true | 3 | [
"bug",
"regression"
] | [] | 2026-04-23T09:32:38Z | 2026-04-28T04:47:41Z | 2026-04-25T19:24:40Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | hellosfs-bot | 251,940,422 | U_kgDODwRORg | User | false |
openclaw/openclaw | 4,314,936,402 | I_kwDOQb6kR88AAAABATC0Ug | 70,555 | https://github.com/openclaw/openclaw/issues/70555 | https://api.github.com/repos/openclaw/openclaw/issues/70555 | [Bug]: Session stuck after gateway restart during tool-use loop — stale lock not recovered | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
After a gateway restart triggered by a config change, a session in an active tool-use loop is not resumed — the session remains permanently stuck in "running" state with tool results persisted but no follow-up ... | closed | completed | true | 2 | [
"bug",
"bug:behavior"
] | [] | 2026-04-23T09:26:54Z | 2026-04-28T04:47:42Z | 2026-04-25T19:24:34Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | ihedy | 1,535,963 | MDQ6VXNlcjE1MzU5NjM= | User | false |
openclaw/openclaw | 4,315,044,546 | I_kwDOQb6kR88AAAABATJawg | 70,559 | https://github.com/openclaw/openclaw/issues/70559 | https://api.github.com/repos/openclaw/openclaw/issues/70559 | runUnsafeReindex crashes with "no such table: chunks_vec" when sqlite-vec is enabled | # `runUnsafeReindex` crashes with "no such table: chunks_vec" when sqlite-vec is enabled
## Environment
- **OpenClaw version:** 2026.4.10 (commit `44e5b62`)
- **Platform:** Windows 11 Pro 10.0.26200
- **Node.js embedded SQLite:** `DatabaseSync` (node:sqlite) with `sqlite-vec` extension
- **Memory backend:** file-back... | open | null | false | 1 | [] | [] | 2026-04-23T09:46:50Z | 2026-05-02T13:08:14Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | Gaia39rus | 272,196,783 | U_kgDOEDlkrw | User | false |
openclaw/openclaw | 4,315,081,810 | I_kwDOQb6kR88AAAABATLsUg | 70,560 | https://github.com/openclaw/openclaw/issues/70560 | https://api.github.com/repos/openclaw/openclaw/issues/70560 | Telegram model picker shows raw IDs instead of display names when navigated via button click | ## Bug: Telegram model picker shows raw IDs instead of display names when navigated via button click
### Summary
When using the Telegram channel and navigating the `/models <provider>` picker by **clicking the provider button**, the model list shows raw model IDs (e.g. `gemini-3.1-pro-preview`) instead of the configur... | closed | completed | true | 0 | [] | [] | 2026-04-23T09:53:34Z | 2026-04-27T04:41:55Z | 2026-04-25T01:38:16Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | MaikiMolto | 150,596,096 | U_kgDOCPnqAA | User | false |
openclaw/openclaw | 4,315,166,273 | I_kwDOQb6kR88AAAABATQ2QQ | 70,563 | https://github.com/openclaw/openclaw/issues/70563 | https://api.github.com/repos/openclaw/openclaw/issues/70563 | Subagent spawn fails: thinking enabled but reasoning_content missing | ## Bug Description
When spawning a subagent via sessions_spawn runtime=subagent, the LLM request is rejected with:
```
FailoverError: LLM request rejected: thinking is enabled but reasoning_content is missing in assistant tool call message at index 2
```
This causes all subagent tasks to fail and fall back to minima... | open | null | false | 1 | [] | [] | 2026-04-23T10:07:46Z | 2026-05-02T21:50:18Z | null | NONE | null | 20260503T000641Z | 2026-05-03T00:06:41Z | Druss520 | 17,692,767 | MDQ6VXNlcjE3NjkyNzY3 | User | false |
openclaw/openclaw | 4,315,216,037 | I_kwDOQb6kR88AAAABATT4pQ | 70,565 | https://github.com/openclaw/openclaw/issues/70565 | https://api.github.com/repos/openclaw/openclaw/issues/70565 | Kimi K2.5 (anthropic-messages) multi-turn tool calling fails: reasoning_content missing in assistant tool_use messages | ## Bug Description
When using **Kimi K2.5** (or any Moonshot model) via the `anthropic-messages` API format with **thinking/reasoning enabled**, multi-turn tool calling fails with a `400 Bad Request` error on the second and subsequent turns.
### Error Message
```
thinking is enabled but reasoning_content is missing ... | closed | completed | true | 0 | [] | [] | 2026-04-23T10:16:31Z | 2026-04-27T05:09:17Z | 2026-04-24T07:13:33Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | Lunareith | 262,595,192 | U_kgDOD6bieA | User | false |
openclaw/openclaw | 4,315,098,183 | I_kwDOQb6kR88AAAABATMsRw | 70,561 | https://github.com/openclaw/openclaw/issues/70561 | https://api.github.com/repos/openclaw/openclaw/issues/70561 | [Bug]: Discord streaming mode leaks tool outputs to users | ## Bug 描述
在 Discord 频道中使用流式输出（`streaming.mode="partial"`）时，工具调用（tool calls）的输出结果被显示给了 Discord 用户，而不是仅供 AI 自己查看。
这些内容本应是内部的，不应该暴露给最终用户。
## 复现步骤
1. 在 Discord 配置中设置 `channels.discord.streaming.mode = "partial"`
2. 让 AI 执行需要工具调用的任务
3. 观察 Discord 频道 —— 工具执行的输出被实时显示出来
## 预期行为
工具输出应该只对 AI 可见，不应该通过流式输出显示给 Discord 用户。
##... | closed | completed | true | 2 | [] | [] | 2026-04-23T09:56:22Z | 2026-04-29T04:36:58Z | 2026-04-26T10:51:28Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | 34262315716 | 62,001,009 | MDQ6VXNlcjYyMDAxMDA5 | User | false |
openclaw/openclaw | 4,314,969,023 | I_kwDOQb6kR88AAAABATEzvw | 70,557 | https://github.com/openclaw/openclaw/issues/70557 | https://api.github.com/repos/openclaw/openclaw/issues/70557 | [Bug]: models.mode=merge may incorrectly drop image attachments when one provider is missing apiKey | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
When models.mode is set to "merge" and multiple providers are configured, image upload may fail unexpectedly if one of the providers is missing the apiKey field in both openclaw.json and agents/main/agent/models.json, even t... | closed | completed | true | 2 | [
"bug",
"regression"
] | [] | 2026-04-23T09:32:58Z | 2026-04-28T04:47:40Z | 2026-04-25T19:24:47Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | LambdaHans | 69,351,383 | MDQ6VXNlcjY5MzUxMzgz | User | false |
openclaw/openclaw | 4,315,353,869 | I_kwDOQb6kR88AAAABATcTDQ | 70,567 | https://github.com/openclaw/openclaw/issues/70567 | https://api.github.com/repos/openclaw/openclaw/issues/70567 | [Bug]: Memory indexer leaves files with Dirty: no after transient /v1/embeddings 500 — vector store silently partial | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
A single transient `500 Internal Server Error` on `/v1/embeddings` during memory indexing left 8 of 10 workspace files unindexed while `openclaw memory status` reported `Dirty: no`, hiding the partial state unt... | closed | completed | true | 1 | [] | [] | 2026-04-23T10:43:19Z | 2026-04-29T04:34:32Z | 2026-04-26T18:57:49Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | chestercs | 18,303,902 | MDQ6VXNlcjE4MzAzOTAy | User | false |
openclaw/openclaw | 4,315,407,942 | I_kwDOQb6kR88AAAABATfmRg | 70,573 | https://github.com/openclaw/openclaw/issues/70573 | https://api.github.com/repos/openclaw/openclaw/issues/70573 | [Bug]: Security Issue: Agent Privacy Isolation Bypass via Direct File Read | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Agents configured with `memorySearch.enabled=false` and separate workspace directories can still access other agent's private memory files using the `read` tool, bypassing all isolation measures.
### Steps to ... | open | null | false | 3 | [
"bug",
"security",
"bug:behavior"
] | [] | 2026-04-23T10:54:06Z | 2026-04-28T15:35:21Z | null | NONE | null | 20260428T180020Z | 2026-04-28T18:00:20Z | iamspark1e | 29,349,655 | MDQ6VXNlcjI5MzQ5NjU1 | User | false |
openclaw/openclaw | 4,315,400,824 | I_kwDOQb6kR88AAAABATfKeA | 70,572 | https://github.com/openclaw/openclaw/issues/70572 | https://api.github.com/repos/openclaw/openclaw/issues/70572 | Session model pin stored without `openrouter/` prefix — bare provider slugs persist across restarts and route to native plugins | ## Summary
The Control UI model dropdown appears to post a **bare provider slug** (e.g. `nvidia/nemotron-3-super-120b-a12b:free`) to the gateway when a user picks a model listed under the OpenRouter catalog. The gateway stores this as-given in `agents/main/sessions/sessions.json`. On subsequent requests, the bare slug... | closed | not_planned | true | 1 | [] | [] | 2026-04-23T10:52:37Z | 2026-04-28T04:47:39Z | 2026-04-25T19:25:00Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | TomCruiseTorpedo | 233,243,487 | U_kgDODecDXw | User | false |
openclaw/openclaw | 4,315,426,000 | I_kwDOQb6kR88AAAABATgs0A | 70,574 | https://github.com/openclaw/openclaw/issues/70574 | https://api.github.com/repos/openclaw/openclaw/issues/70574 | [Bug]: Subagent completion announce routes to child agent's channel binding instead of parent agent's session | ## Bug Description
When Agent A (parent) spawns Agent B (child) via `sessions_spawn`, Agent B's completion announce should route back to Agent A's session context. Instead, the current implementation overrides Agent A's `accountId` with Agent B's binding `accountId`. At completion time, the bound delivery router finds... | closed | completed | true | 2 | [] | [] | 2026-04-23T10:57:33Z | 2026-04-30T04:34:00Z | 2026-04-27T13:18:06Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | mlaihk | 25,972,362 | MDQ6VXNlcjI1OTcyMzYy | User | false |
openclaw/openclaw | 4,315,447,240 | I_kwDOQb6kR88AAAABATh_yA | 70,575 | https://github.com/openclaw/openclaw/issues/70575 | https://api.github.com/repos/openclaw/openclaw/issues/70575 | Stale-entry writeback races for lastChannel/channel in agent handler session patch | ## Context
PR #19328 (fix for #5369) removed most stale-entry writebacks from the session-entry patch in `src/gateway/server-methods/agent.ts`. Two residual patterns in the same function still carry a stale cached value into the patch and can clobber a fresh concurrent write via the `{...freshStore, ...patch}` merge i... | open | null | false | 1 | [] | [] | 2026-04-23T11:01:17Z | 2026-05-02T13:08:16Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | CodeReclaimers | 12,574,428 | MDQ6VXNlcjEyNTc0NDI4 | User | false |
openclaw/openclaw | 4,315,468,130 | I_kwDOQb6kR88AAAABATjRYg | 70,576 | https://github.com/openclaw/openclaw/issues/70576 | https://api.github.com/repos/openclaw/openclaw/issues/70576 | Feature: Forward reply_to_message context in Telegram inbound metadata | ## Problem
When a user replies to a specific message in a Telegram chat, the OpenClaw agent does not receive the referenced message content. The Telegram Bot API provides `reply_to_message` in the update payload, but this information is not forwarded to the agent in the inbound metadata.
This means:
- Agents cannot un... | closed | completed | true | 1 | [] | [] | 2026-04-23T11:05:29Z | 2026-04-28T04:47:38Z | 2026-04-25T19:25:07Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | garlem42-dev | 278,662,299 | U_kgDOEJwMmw | User | false |
openclaw/openclaw | 4,315,468,930 | I_kwDOQb6kR88AAAABATjUgg | 70,577 | https://github.com/openclaw/openclaw/issues/70577 | https://api.github.com/repos/openclaw/openclaw/issues/70577 | memory_search tool can fail with CredentialsProviderError while CLI works after switching to local embeddings | ## Summary
In a real local install, the built-in `memory_search` tool kept failing with:
```text
Could not load credentials from any providers
```
while the underlying memory store was present and `openclaw memory status --deep` reported a configured provider. I was able to repair the CLI/runtime by switching `agent... | closed | not_planned | true | 2 | [] | [] | 2026-04-23T11:05:39Z | 2026-04-30T04:38:45Z | 2026-04-26T05:27:29Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | malshaalan-ai | 260,099,975 | U_kgDOD4DPhw | User | false |
openclaw/openclaw | 4,315,498,319 | I_kwDOQb6kR88AAAABATlHTw | 70,578 | https://github.com/openclaw/openclaw/issues/70578 | https://api.github.com/repos/openclaw/openclaw/issues/70578 | Silent config data loss: Zod `.strict()` strips unknown top-level fields on round-trip | # Silent config data loss: Zod `.strict()` + `projectSourceOntoRuntimeShape` + CLI-surface allowlist design
## TL;DR
On OpenClaw 2026.4.15 (Windows, global npm install), my `openclaw.json` has been silently stripped from ~34 KB to ~13 KB on five occasions since 2026-03-22, triggered by `openclaw plugins install <path... | open | null | false | 1 | [] | [] | 2026-04-23T11:11:11Z | 2026-05-02T12:54:09Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | grimmjoww | 20,139,622 | MDQ6VXNlcjIwMTM5NjIy | User | false |
openclaw/openclaw | 4,315,503,496 | I_kwDOQb6kR88AAAABATlbiA | 70,580 | https://github.com/openclaw/openclaw/issues/70580 | https://api.github.com/repos/openclaw/openclaw/issues/70580 | Media preflight: carry transcript state into downstream media-understanding to avoid duplicate audio STT | ## Summary
Channel-level audio preflight helpers can transcribe an inbound voice note before the normal reply pipeline runs, but that transcript state is not carried into downstream `applyMediaUnderstanding` / attachment selection. As a result, the same inbound audio can be transcribed again on the normal media-unders... | closed | completed | true | 2 | [] | [] | 2026-04-23T11:12:13Z | 2026-04-30T04:42:06Z | 2026-04-26T04:49:09Z | CONTRIBUTOR | null | 20260430T060018Z | 2026-04-30T06:00:18Z | rogerdigital | 13,251,150 | MDQ6VXNlcjEzMjUxMTUw | User | false |
openclaw/openclaw | 4,315,540,027 | I_kwDOQb6kR88AAAABATnqOw | 70,581 | https://github.com/openclaw/openclaw/issues/70581 | https://api.github.com/repos/openclaw/openclaw/issues/70581 | [Bug]: macOS LaunchAgent gateway cannot reliably use env-backed model-provider SecretRefs even when local CLI can | ## Summary
On macOS LaunchAgent installs, gateway-managed inference cannot reliably use an env-backed model-provider API key even when the exact same provider config works for local CLI inference.
Concrete example:
```json
{
"models": {
"providers": {
"openrouter": {
"baseUrl": "https://openroute... | closed | completed | true | 1 | [] | [] | 2026-04-23T11:19:07Z | 2026-04-28T04:47:37Z | 2026-04-25T19:25:14Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | VishalJ99 | 51,826,812 | MDQ6VXNlcjUxODI2ODEy | User | false |
openclaw/openclaw | 4,315,540,957 | I_kwDOQb6kR88AAAABATnt3Q | 70,582 | https://github.com/openclaw/openclaw/issues/70582 | https://api.github.com/repos/openclaw/openclaw/issues/70582 | LCM compaction sends deprecated parameter to claude-opus-4-7, causing repeated API errors and retry loops | ## Bug Report
### Summary
After upgrading to OpenClaw 2026.4.21, LCM (Lossless Context Management) compaction is sending `temperature: 0.2` to `anthropic/claude-opus-4-7`, which no longer accepts this parameter. This causes every compaction attempt to fail with a 400 error and then retry, wasting tokens and introducin... | closed | completed | true | 4 | [] | [] | 2026-04-23T11:19:17Z | 2026-04-29T04:36:43Z | 2026-04-26T07:16:22Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | Joe2321 | 129,733,981 | U_kgDOB7uVXQ | User | false |
openclaw/openclaw | 4,222,056,025 | I_kwDOQb6kR877p3ZZ | 62,919 | https://github.com/openclaw/openclaw/issues/62919 | https://api.github.com/repos/openclaw/openclaw/issues/62919 | 2026.4.7: Telegram extension dist missing channel.setup.js — startup crash | ## Summary
OpenClaw 2026.4.7 ships with a missing file in the Telegram extension dist bundle:
```
~/.nvm/versions/node/v24.14.0/lib/node_modules/openclaw/dist/extensions/telegram/src/channel.setup.js
```
This causes a hard startup crash on every CLI command, including `openclaw doctor --fix`.
## Impact
Complete ga... | closed | completed | true | 2 | [] | [] | 2026-04-08T04:14:16Z | 2026-04-27T05:09:29Z | 2026-04-24T05:48:56Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | brobertson07 | 1,428,100 | MDQ6VXNlcjE0MjgxMDA= | User | false |
openclaw/openclaw | 4,315,639,513 | I_kwDOQb6kR88AAAABATtu2Q | 70,587 | https://github.com/openclaw/openclaw/issues/70587 | https://api.github.com/repos/openclaw/openclaw/issues/70587 | [Bug]: Cannot find module '@larksuiteoapi/node-sdk' | ### Bug type
Crash (process/app exits or hangs)
### Beta release blocker
No
### Summary
Trying to install Openclaw on ubuntu (actuall version: v2026.4.21) then crash because of modules not founded - Start by feishu then others following with same error message "can not find module"
### Steps to reproduce
curl -f... | closed | completed | true | 7 | [
"bug"
] | [] | 2026-04-23T11:39:12Z | 2026-04-26T04:30:13Z | 2026-04-23T16:44:06Z | NONE | null | 20260426T060034Z | 2026-04-26T06:00:34Z | Everyways | 6,754,974 | MDQ6VXNlcjY3NTQ5NzQ= | User | false |
openclaw/openclaw | 4,315,650,209 | I_kwDOQb6kR88AAAABATuYoQ | 70,588 | https://github.com/openclaw/openclaw/issues/70588 | https://api.github.com/repos/openclaw/openclaw/issues/70588 | Feature Request: Add webhook-create action to Discord message tool | **Is your feature request related to a problem? Please describe.**
Currently, the `message` tool for the Discord channel plugin allows creating channels, categories, and threads, but it lacks an action to create webhooks. Even when the bot role is granted the "Manage Webhooks" permission in Discord, agents cannot progr... | closed | not_planned | true | 1 | [] | [] | 2026-04-23T11:41:25Z | 2026-04-29T04:34:00Z | 2026-04-26T21:31:42Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | Stache73 | 7,196,384 | MDQ6VXNlcjcxOTYzODQ= | User | false |
openclaw/openclaw | 4,315,720,163 | I_kwDOQb6kR88AAAABATyp4w | 70,590 | https://github.com/openclaw/openclaw/issues/70590 | https://api.github.com/repos/openclaw/openclaw/issues/70590 | [Feature]: Solve the problem of hard-coded language | ### Summary
Hello, please note that many languages in the webUI are still in English. Please improve the multilingual support. Thank you
### Problem to solve
It is inconvenient for users of other languages to use
### Proposed solution
Translate the residual and hard-coded content into multiple languages
### Alt... | closed | not_planned | true | 1 | [
"enhancement"
] | [] | 2026-04-23T11:54:42Z | 2026-04-29T04:33:38Z | 2026-04-26T22:00:55Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | fangguisheng520 | 237,461,326 | U_kgDODidfTg | User | false |
openclaw/openclaw | 4,315,775,561 | I_kwDOQb6kR88AAAABAT2CSQ | 70,591 | https://github.com/openclaw/openclaw/issues/70591 | https://api.github.com/repos/openclaw/openclaw/issues/70591 | [ClawX] Session freezes after compaction — Thinking... stuck, stop button not working | ## Bug Description
After ClawX 3.10 + OpenClaw 2026.4.21 upgrade, sessions freeze in 'Thinking...' state after compaction triggers. The bot reacts with 👀 emoji but the LLM call never completes. Stop button does not work.
**This appears to be the same root cause as #60867 (Discord) but in ClawX webchat.**
## Steps t... | closed | not_planned | true | 1 | [] | [] | 2026-04-23T12:04:48Z | 2026-04-29T04:42:02Z | 2026-04-26T07:16:36Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | TRDEFI | 223,904,629 | U_kgDODViDdQ | User | false |
openclaw/openclaw | 4,204,901,721 | I_kwDOQb6kR876obVZ | 60,867 | https://github.com/openclaw/openclaw/issues/60867 | https://api.github.com/repos/openclaw/openclaw/issues/60867 | Discord group session freezes after compaction — reacts with 👀 but never responds | ## Bug Description
Discord group chat sessions become unresponsive after context compaction when the session reaches ~100K-130K tokens. The bot acknowledges incoming messages (adds 👀 reaction) but never generates a response. The session appears alive (not timed out, not errored) but is permanently stuck.
## Steps to... | closed | completed | true | 2 | [] | [] | 2026-04-04T13:07:00Z | 2026-04-28T05:10:20Z | 2026-04-25T13:46:51Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | altanests | 201,349,295 | U_kgDODABYrw | User | false |
openclaw/openclaw | 4,315,800,417 | I_kwDOQb6kR88AAAABAT3jYQ | 70,594 | https://github.com/openclaw/openclaw/issues/70594 | https://api.github.com/repos/openclaw/openclaw/issues/70594 | [Bug] channel status command hangs indefinitely on openclaw-weixin | **Description:** When running openclaw channel status openclaw-weixin, the command hangs indefinitely and never returns.
**Environment:**
- QClaw version: latest
- Windows 10
- Channel: openclaw-weixin (WeChat)
**Steps to reproduce:**
1. Run: openclaw channel status openclaw-weixin
2. Command hangs, no output, no err... | closed | completed | true | 2 | [] | [] | 2026-04-23T12:09:15Z | 2026-04-28T04:47:36Z | 2026-04-25T19:25:20Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | 15285701683 | 235,316,189 | U_kgDODgaj3Q | User | false |
openclaw/openclaw | 4,315,848,840 | I_kwDOQb6kR88AAAABAT6giA | 70,600 | https://github.com/openclaw/openclaw/issues/70600 | https://api.github.com/repos/openclaw/openclaw/issues/70600 | [Bug]: docs say 'openclaw devices approve --latest' auto-approves, but 2026.4.15 CLI is preview-only | ## Summary
On OpenClaw 2026.4.15, `openclaw devices approve --latest` is preview-only and exits non-zero instead of approving the latest pending request, but the public docs still say it auto-approves when `requestId` is omitted or `--latest` is passed.
This creates a real operator trap during device pairing because ... | closed | not_planned | true | 1 | [] | [] | 2026-04-23T12:17:57Z | 2026-04-29T04:36:22Z | 2026-04-26T12:06:18Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | wbyanclaw | 271,781,585 | U_kgDOEDMO0Q | User | false |
openclaw/openclaw | 4,315,898,632 | I_kwDOQb6kR88AAAABAT9jCA | 70,603 | https://github.com/openclaw/openclaw/issues/70603 | https://api.github.com/repos/openclaw/openclaw/issues/70603 | Difficulty Spike in stage_004 (ROOT ACCESS) causing significant loot loss | Confirmed by multiple agents (Voss, Ghost). Agents report a substantial difficulty jump that leads to unsustainable losses of high-value loot and makes the deployment process feel like a gamble rather than strategy. | closed | not_planned | true | 1 | [] | [] | 2026-04-23T12:26:18Z | 2026-04-28T04:47:36Z | 2026-04-25T19:25:34Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | mons-bot | 262,623,366 | U_kgDOD6dQhg | User | false |
openclaw/openclaw | 4,079,365,305 | I_kwDOQb6kR87zJiy5 | 47,668 | https://github.com/openclaw/openclaw/issues/47668 | https://api.github.com/repos/openclaw/openclaw/issues/47668 | Context/memory accumulates too quickly causing frequent session resets | Session context fills up within ~80-100 messages, requiring frequent /new resets. This disrupts conversation continuity and feels jarring for the user.
**Expected behavior:** Sessions should handle longer conversations gracefully, or have a smoother checkpoint/resume mechanism.
**Current workaround:** Users resort to... | open | null | false | 5 | [] | [] | 2026-03-15T22:43:53Z | 2026-05-04T01:22:25Z | null | NONE | null | 20260504T030031Z | 2026-05-04T03:00:31Z | Jragyn | 7,897,461 | MDQ6VXNlcjc4OTc0NjE= | User | false |
openclaw/openclaw | 4,315,924,304 | I_kwDOQb6kR88AAAABAT_HUA | 70,604 | https://github.com/openclaw/openclaw/issues/70604 | https://api.github.com/repos/openclaw/openclaw/issues/70604 | [Bug]: Image tool ignores user-configured provider baseUrl for zai/glm-5v-turbo, routes to vercel-ai-gateway instead (404) | ## Bug type
Behavior bug (incorrect output/state without crash)
## Beta release blocker
No
## Summary
The `image` tool always routes `zai/glm-5v-turbo` through pi-ai's built-in model registry entry (`vercel-ai-gateway` + `anthropic-messages` API) instead of the user-configured `zai` provider (`open.bigmodel.cn/api/co... | closed | completed | true | 1 | [] | [] | 2026-04-23T12:30:47Z | 2026-04-29T04:42:02Z | 2026-04-26T07:16:41Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | Sway-Chan | 15,033,876 | MDQ6VXNlcjE1MDMzODc2 | User | false |
openclaw/openclaw | 4,315,940,065 | I_kwDOQb6kR88AAAABAUAE4Q | 70,606 | https://github.com/openclaw/openclaw/issues/70606 | https://api.github.com/repos/openclaw/openclaw/issues/70606 | [Feature]: Batch API support for async-tolerant cron jobs (50% token discount) | ### Summary
Add support for Anthropic's Message Batches API (POST /v1/messages/batches) as a delivery path for cron jobs whose output isn't latency-sensitive. Batch API offers a flat 50% discount on input + output tokens, with up to 24h completion — most complete in under 1h. Natural fit for digest/report jobs that fi... | closed | not_planned | true | 1 | [
"enhancement"
] | [] | 2026-04-23T12:33:46Z | 2026-04-29T04:42:01Z | 2026-04-26T07:16:46Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | htuyer121 | 277,202,213 | U_kgDOEIXFJQ | User | false |
openclaw/openclaw | 4,315,969,926 | I_kwDOQb6kR88AAAABAUB5hg | 70,608 | https://github.com/openclaw/openclaw/issues/70608 | https://api.github.com/repos/openclaw/openclaw/issues/70608 | [Feature]: WhatsApp web idle watchdog should be configurable (30m no-message reconnect) | ### Summary
```
### Summary
The WhatsApp web runtime currently forces a reconnect after ~30 minutes with no inbound messages:
`No messages received in 30m - restarting connection`
This seems to be a fixed idle watchdog, not a user-configurable setting.
### Environment
- OpenClaw 2026.4.21
- WhatsApp web runtime
-... | closed | not_planned | true | 1 | [
"enhancement"
] | [] | 2026-04-23T12:39:15Z | 2026-04-29T04:42:00Z | 2026-04-26T07:16:50Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | samircaki | 278,689,085 | U_kgDOEJx1PQ | User | false |
openclaw/openclaw | 4,315,976,050 | I_kwDOQb6kR88AAAABAUCRcg | 70,610 | https://github.com/openclaw/openclaw/issues/70610 | https://api.github.com/repos/openclaw/openclaw/issues/70610 | Control UI agent switcher shows raw session key instead of agent display name for non-default agents | ## Description
In Control UI (left sidebar agent switcher), non-default agents display the raw internal session key format instead of the configured display name.
**Expected:** `旺财(lottery)/main`
**Actual:** `webchat:g-agent-lottery-main`
## Environment
- OpenClaw version: 2026.4.21 (f788c88)
- OS: macOS (arm64, Da... | closed | completed | true | 1 | [] | [] | 2026-04-23T12:40:19Z | 2026-04-28T04:47:35Z | 2026-04-25T19:25:40Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | antaoaee | 275,708,440 | U_kgDOEG76GA | User | false |
openclaw/openclaw | 4,316,118,163 | I_kwDOQb6kR88AAAABAUK8kw | 70,612 | https://github.com/openclaw/openclaw/issues/70612 | https://api.github.com/repos/openclaw/openclaw/issues/70612 | [Bug]: gateway install --force does not remove embedded service token in 2026.4.21 | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Version: 2026.4.21 (f788c88)
Steps: run openclaw gateway install --force
Expected: plist contains no literal secret <string> values, only OPENCLAW_SERVICE_MANAGED_ENV_KEYS passthrough
Actual: plist contains both — passthroug... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-23T13:04:40Z | 2026-04-28T04:47:34Z | 2026-04-25T19:25:47Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Phluttr | 268,270,571 | U_kgDOD_176w | User | false |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.