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 | 4,252,851,872 | I_kwDOQb6kR879fV6g | 65,817 | https://github.com/openclaw/openclaw/issues/65817 | https://api.github.com/repos/openclaw/openclaw/issues/65817 | [Feature] Sub-agent cascade circuit breaker | ## Problem
When a sub-agent fails (bad config, missing dependency, logic error), there is no mechanism to prevent the parent from immediately spawning it again. This creates cascade failures where the same broken sub-agent is retried indefinitely, burning tokens and time.
OpenClaw tracks sub-agent completion status a... | closed | not_planned | true | 1 | [] | [] | 2026-04-13T08:50:04Z | 2026-04-21T04:21:32Z | 2026-04-18T23:58:04Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | smonett | 50,761,481 | MDQ6VXNlcjUwNzYxNDgx | User | false |
openclaw/openclaw | 4,252,854,108 | I_kwDOQb6kR879fWdc | 65,818 | https://github.com/openclaw/openclaw/issues/65818 | https://api.github.com/repos/openclaw/openclaw/issues/65818 | [Feature] Doctor dry-run / diff mode | ## Problem
`openclaw doctor --fix` and `--repair` apply changes to the config without showing what will change beforehand. While `--non-interactive` limits to safe migrations, there is no way to preview the full set of changes doctor would make before committing them.
This is risky because doctor performs both migrat... | closed | not_planned | true | 1 | [] | [] | 2026-04-13T08:50:19Z | 2026-04-21T04:21:32Z | 2026-04-18T23:58:06Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | smonett | 50,761,481 | MDQ6VXNlcjUwNzYxNDgx | User | false |
openclaw/openclaw | 4,252,856,057 | I_kwDOQb6kR879fW75 | 65,819 | https://github.com/openclaw/openclaw/issues/65819 | https://api.github.com/repos/openclaw/openclaw/issues/65819 | [Feature] Post-update hook system | ## Problem
`openclaw update` fetches the latest version, runs doctor, and restarts. Any local dist patches (hotfixes for upstream bugs applied to files in the OpenClaw install directory) are overwritten with no way to auto-reapply them.
Users who patch upstream bugs locally between releases lose their fixes on every ... | closed | not_planned | true | 1 | [] | [] | 2026-04-13T08:50:32Z | 2026-04-21T04:21:31Z | 2026-04-18T23:58:09Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | smonett | 50,761,481 | MDQ6VXNlcjUwNzYxNDgx | User | false |
openclaw/openclaw | 4,252,857,783 | I_kwDOQb6kR879fXW3 | 65,820 | https://github.com/openclaw/openclaw/issues/65820 | https://api.github.com/repos/openclaw/openclaw/issues/65820 | [Feature] MCP tool file access deny list | ## Problem
Native Read/Write tools respect file deny lists. MCP tools that perform file operations (e.g., `mcp__filesystem__read_file`) use their own access paths and may bypass the native deny list, allowing MCP servers to read credentials, .env files, and other sensitive data.
## Workaround
I built `cc-mcp-guard.p... | closed | not_planned | true | 1 | [] | [] | 2026-04-13T08:50:43Z | 2026-04-21T04:21:30Z | 2026-04-18T23:58:12Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | smonett | 50,761,481 | MDQ6VXNlcjUwNzYxNDgx | User | false |
openclaw/openclaw | 4,252,859,666 | I_kwDOQb6kR879fX0S | 65,821 | https://github.com/openclaw/openclaw/issues/65821 | https://api.github.com/repos/openclaw/openclaw/issues/65821 | [Feature] Quality gate hooks on agent file writes | ## Problem
When agents write or edit files (via Write/Edit tools), there is no automated quality check. Agents can introduce syntax errors, security vulnerabilities, or lint violations without immediate feedback.
## Workaround
I built `cc-quality-gate.py`, a PostToolUse hook that runs `ruff` (lint) + `bandit` (secur... | closed | not_planned | true | 1 | [] | [] | 2026-04-13T08:50:57Z | 2026-04-21T04:21:30Z | 2026-04-18T23:58:15Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | smonett | 50,761,481 | MDQ6VXNlcjUwNzYxNDgx | User | false |
openclaw/openclaw | 4,252,861,653 | I_kwDOQb6kR879fYTV | 65,822 | https://github.com/openclaw/openclaw/issues/65822 | https://api.github.com/repos/openclaw/openclaw/issues/65822 | [Feature] Per-binary exec trust (path or hash) | ## Problem
The current exec trust model uses `safeBinTrustedDirs` — entire directories whose binaries are trusted. This is coarse: trusting `/opt/homebrew/bin` trusts every binary Homebrew installs, including potentially untrusted packages.
The alternative is creating individual wrapper scripts in a trusted directory... | closed | not_planned | true | 1 | [] | [] | 2026-04-13T08:51:10Z | 2026-04-21T04:21:29Z | 2026-04-18T23:58:18Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | smonett | 50,761,481 | MDQ6VXNlcjUwNzYxNDgx | User | false |
openclaw/openclaw | 4,252,863,660 | I_kwDOQb6kR879fYys | 65,823 | https://github.com/openclaw/openclaw/issues/65823 | https://api.github.com/repos/openclaw/openclaw/issues/65823 | [Feature] Multi-agent output provenance tracking | ## Problem
When multiple agents or models contribute to a synthesis (e.g., one model writes, another reviews, a third judges), the final output may claim multi-model validation. There is no native mechanism to verify that a "multi-model synthesis" actually involved multiple models — an agent could claim consensus with... | closed | not_planned | true | 1 | [] | [] | 2026-04-13T08:51:24Z | 2026-04-21T04:21:28Z | 2026-04-18T23:58:21Z | NONE | null | 20260421T060038Z | 2026-04-21T06:00:38Z | smonett | 50,761,481 | MDQ6VXNlcjUwNzYxNDgx | User | false |
openclaw/openclaw | 4,252,870,394 | I_kwDOQb6kR879fab6 | 65,824 | https://github.com/openclaw/openclaw/issues/65824 | https://api.github.com/repos/openclaw/openclaw/issues/65824 | [Meta] Feature request bundle: 11 platform gaps from intensive daily use | ## OpenClaw Feature Requests — Consolidated & Audited
**Author:** @smonett (intensive daily user since 2026.2)
**Audited against:** OpenClaw 2026.4.15 (April 17, 2026) · Refined 2026-04-19 after deeper package audit
**Context:** Every item below is backed by a working workaround script we actively run in production. H... | open | null | false | 13 | [] | [] | 2026-04-13T08:52:09Z | 2026-05-03T06:07:22Z | null | NONE | null | 20260503T090142Z | 2026-05-03T09:01:42Z | smonett | 50,761,481 | MDQ6VXNlcjUwNzYxNDgx | User | false |
openclaw/openclaw | 4,252,996,153 | I_kwDOQb6kR879f5I5 | 65,829 | https://github.com/openclaw/openclaw/issues/65829 | https://api.github.com/repos/openclaw/openclaw/issues/65829 | [Bug]: image tool fails with custom provider — ModelRegistry doesn't recognize custom providers from models.json | ## Description
The `image` tool fails when `imageModel` is set to a custom provider because the tool internally uses `@mariozechner/pi-coding-agent`'s `ModelRegistry.find()`, which does not properly load custom providers from `models.json`.
## Error
```
Unknown model: custom-provider/model-id
```
## Root Cause
Two... | closed | completed | true | 0 | [] | [] | 2026-04-13T09:07:04Z | 2026-04-16T04:24:43Z | 2026-04-13T09:14:31Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | GentleCold | 60,008,333 | MDQ6VXNlcjYwMDA4MzMz | User | false |
openclaw/openclaw | 4,253,049,613 | I_kwDOQb6kR879gGMN | 65,832 | https://github.com/openclaw/openclaw/issues/65832 | https://api.github.com/repos/openclaw/openclaw/issues/65832 | [Bug]: image tool fails with custom provider — ModelRegistry doesn't recognize custom providers from models.json | ## Bug type
Behavior bug (incorrect output/state without crash)
## Summary
The `image` tool throws `Unknown model: <custom-provider>/<model-id>` when `imageModel` is configured with a custom provider defined in `models.json`, because the tool internally uses `@mariozechner/pi-coding-agent`'s `ModelRegistry.find()` w... | closed | not_planned | true | 1 | [
"clownfish"
] | [] | 2026-04-13T09:15:11Z | 2026-04-28T04:32:16Z | 2026-04-25T19:57:58Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | GentleCold | 60,008,333 | MDQ6VXNlcjYwMDA4MzMz | User | false |
openclaw/openclaw | 4,253,077,500 | I_kwDOQb6kR879gM_8 | 65,834 | https://github.com/openclaw/openclaw/issues/65834 | https://api.github.com/repos/openclaw/openclaw/issues/65834 | [Bug]: memory-core dreaming cron reconciliation can report "cron service unavailable" even when gateway cron is running | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
`memory-core` sometimes logs:
`memory-core: managed dreaming cron could not be reconciled (cron service unavailable).`
But gateway cron is actually enabled and running.
This looks like a runtime bug in how `... | closed | completed | true | 6 | [
"bug"
] | [] | 2026-04-13T09:19:48Z | 2026-04-30T04:40:58Z | 2026-04-26T05:00:44Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | vavrusa78 | 108,886,693 | U_kgDOBn16pQ | User | false |
openclaw/openclaw | 4,253,139,828 | I_kwDOQb6kR879gcN0 | 65,835 | https://github.com/openclaw/openclaw/issues/65835 | https://api.github.com/repos/openclaw/openclaw/issues/65835 | sourceConfig in config.get exposes API keys in plaintext | ## Describe the bug
When calling the \gateway config.get\ tool, the \sourceConfig\ field in the response returns the raw contents of the config file with **all API keys in plaintext** (e.g. \piKey\, \ppSecret\, \ oken\).
Meanwhile, the \parsed\ and \
esolved\ fields in the same response are properly redacted (showi... | closed | completed | true | 2 | [] | [] | 2026-04-13T09:26:53Z | 2026-04-17T04:21:40Z | 2026-04-15T03:51:34Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | JIRBOY | 49,422,297 | MDQ6VXNlcjQ5NDIyMjk3 | User | false |
openclaw/openclaw | 4,253,182,071 | I_kwDOQb6kR879gmh3 | 65,836 | https://github.com/openclaw/openclaw/issues/65836 | https://api.github.com/repos/openclaw/openclaw/issues/65836 | sourceConfig in config.get exposes API keys in plaintext | ## Describe the bug
When calling the \gateway config.get\ tool, the \sourceConfig\ field in the response returns the raw contents of the config file with **all API keys in plaintext** (e.g. \piKey\, \ppSecret\, \ oken\).
Meanwhile, the \parsed\ and \
esolved\ fields in the same response are properly redacted (showi... | closed | completed | true | 1 | [] | [] | 2026-04-13T09:30:38Z | 2026-04-16T04:24:42Z | 2026-04-13T09:30:50Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | JIRBOY | 49,422,297 | MDQ6VXNlcjQ5NDIyMjk3 | User | false |
openclaw/openclaw | 4,253,242,058 | I_kwDOQb6kR879g1LK | 65,838 | https://github.com/openclaw/openclaw/issues/65838 | https://api.github.com/repos/openclaw/openclaw/issues/65838 | [Feature Request] Add option to disable Bonjour/mDNS advertiser on Windows | ## Problem
On Windows systems without Apple Bonjour service installed, the OpenClaw gateway startup is slow (~50 seconds) because the mDNS advertiser gets stuck trying to broadcast:
```
[bonjour] watchdog detected non-announced service; attempting re-advertise
[bonjour] restarting advertiser (service stuck in announc... | closed | completed | true | 2 | [] | [] | 2026-04-13T09:36:40Z | 2026-04-28T04:56:44Z | 2026-04-25T17:27:48Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Fission21 | 80,576,421 | MDQ6VXNlcjgwNTc2NDIx | User | false |
openclaw/openclaw | 4,253,250,179 | I_kwDOQb6kR879g3KD | 65,839 | https://github.com/openclaw/openclaw/issues/65839 | https://api.github.com/repos/openclaw/openclaw/issues/65839 | Context limit exceeded error suggests too low reserveTokensFloor value (20000) | ## Problem
When the context limit is exceeded, the error message suggests:
> To prevent this, increase your compaction buffer by setting agents.defaults.compaction.reserveTokensFloor to 20000 or higher in your config.
However, 20000 appears to be too low and users still experience context limit issues even after sett... | open | null | false | 1 | [] | [] | 2026-04-13T09:37:32Z | 2026-05-03T04:43:32Z | null | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | matingxiu | 24,728,784 | MDQ6VXNlcjI0NzI4Nzg0 | User | false |
openclaw/openclaw | 4,253,264,131 | I_kwDOQb6kR879g6kD | 65,840 | https://github.com/openclaw/openclaw/issues/65840 | https://api.github.com/repos/openclaw/openclaw/issues/65840 | [Bug]: image tool reports "Unknown model" for registered imageModel (qwen/qwen3.6-plus) | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Bug 描述
openclaw models list 正确注册了 qwen/qwen3.6-plus 并标记了 image tag，但 image 工具调用时报错 Unknown model: qwen/qwen3.6-plus。
环境
OpenClaw: 2026.4.11 (769908e)
OS: Linux 6.17.0-20-generic (x64)
Node: v22.22.2
配置
... | closed | completed | true | 1 | [
"bug",
"bug:behavior"
] | [] | 2026-04-13T09:39:27Z | 2026-04-28T04:56:43Z | 2026-04-25T17:27:54Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | itanyplus | 34,127,603 | MDQ6VXNlcjM0MTI3NjAz | User | false |
openclaw/openclaw | 4,253,311,088 | I_kwDOQb6kR879hGBw | 65,843 | https://github.com/openclaw/openclaw/issues/65843 | https://api.github.com/repos/openclaw/openclaw/issues/65843 | [Bug]: Paring failure when add remote OpenClaw as ACP server | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
I've set up one OpenClaw on Mac (x86) and another OpenClaw on an Azure VM. I want to link the OpenClaw agent from Mac as an ACP client to the Azure VM as an ACP server. The OpenClaw on the Azure VM is reverse-p... | closed | not_planned | true | 3 | [
"bug",
"bug:behavior"
] | [] | 2026-04-13T09:47:10Z | 2026-04-28T04:56:42Z | 2026-04-25T17:28:01Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | HaoHoo | 10,110,568 | MDQ6VXNlcjEwMTEwNTY4 | User | false |
openclaw/openclaw | 4,253,324,300 | I_kwDOQb6kR879hJQM | 65,844 | https://github.com/openclaw/openclaw/issues/65844 | https://api.github.com/repos/openclaw/openclaw/issues/65844 | [Bug]: Codex multi-profile auth collapsed to openai-codex:default on 2026.4.11, causing auth.order to exclude the only remaining profile | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
After updating, my existing multi-profile OpenAI Codex auth setup appears to have collapsed to a single profile (`openai-codex:default`).
I previously had 5 Codex profiles and a matching global `auth.order.openai-codex` in ... | closed | completed | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-13T09:49:15Z | 2026-04-28T04:56:41Z | 2026-04-25T17:28:07Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | PattrnData | 187,018,763 | U_kgDOCyWuCw | User | false |
openclaw/openclaw | 4,253,382,988 | I_kwDOQb6kR879hXlM | 65,846 | https://github.com/openclaw/openclaw/issues/65846 | https://api.github.com/repos/openclaw/openclaw/issues/65846 | [Bug]: Agent process does not exit after task completion when used via CLI | ## Bug Description
When running openclaw via CLI with \`openclaw agent --local --json\`, the agent process does not exit after completing a task. It remains in a \`livenessState: "working"\` state indefinitely until externally killed or timeout.
## Environment
- OpenClaw version: 2026.4.11
- Platform: Linux (Ubuntu)... | closed | completed | true | 1 | [] | [] | 2026-04-13T09:58:26Z | 2026-04-28T04:56:40Z | 2026-04-25T17:28:14Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | leaderlemon | 97,303,244 | U_kgDOBcy6zA | User | false |
openclaw/openclaw | 4,253,385,060 | I_kwDOQb6kR879hYFk | 65,847 | https://github.com/openclaw/openclaw/issues/65847 | https://api.github.com/repos/openclaw/openclaw/issues/65847 | Bug: agents.defaults.heartbeat.to corrupts deliveryContext, causing private messages to leak to group with different content | ## Summary
When `agents.defaults.heartbeat.to` is configured, **all outgoing messages** (including normal replies to user DMs) get incorrectly routed to the heartbeat target channel. Moreover, the content sent to the wrong channel is **different** from the content sent to the correct channel — as if two independent LL... | closed | completed | true | 3 | [] | [] | 2026-04-13T09:58:45Z | 2026-04-28T04:56:39Z | 2026-04-25T17:28:20Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | smallmj | 23,742,160 | MDQ6VXNlcjIzNzQyMTYw | User | false |
openclaw/openclaw | 4,253,388,645 | I_kwDOQb6kR879hY9l | 65,848 | https://github.com/openclaw/openclaw/issues/65848 | https://api.github.com/repos/openclaw/openclaw/issues/65848 | [Bug]: Can not install skills with python script | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
I am running v2026.4.5 version . Deploy openclaw in docker .
when i in web ui ask openclaw to use clawhub search and install a skill which is depend on installing pip library. It will always failed for first ... | closed | completed | true | 1 | [
"bug",
"bug:behavior"
] | [] | 2026-04-13T09:59:20Z | 2026-04-16T04:24:36Z | 2026-04-13T13:51:20Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | jimmylang74 | 47,094,833 | MDQ6VXNlcjQ3MDk0ODMz | User | false |
openclaw/openclaw | 4,253,397,765 | I_kwDOQb6kR879hbMF | 65,849 | https://github.com/openclaw/openclaw/issues/65849 | https://api.github.com/repos/openclaw/openclaw/issues/65849 | message tool: Discord send fails unless components.modal is populated (should be optional) | ## Bug Description
When using the unified `message` tool with `action=send` to Discord, plain text sends fail validation unless a `components.modal` structure is populated.
In our setup, attempts to send a normal message to a Discord channel fail with errors such as:
- `components.modal.fields must be a non-empty arra... | closed | completed | true | 1 | [] | [] | 2026-04-13T10:00:47Z | 2026-04-28T04:56:38Z | 2026-04-25T17:28:27Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Skrblik | 139,452,656 | U_kgDOCE_g8A | User | false |
openclaw/openclaw | 4,253,404,268 | I_kwDOQb6kR879hcxs | 65,851 | https://github.com/openclaw/openclaw/issues/65851 | https://api.github.com/repos/openclaw/openclaw/issues/65851 | qqbot missing from channel contract guardrails and package-manifest checks | ## Description
`qqbot` is a bundled channel plugin (`extensions/qqbot/`) with three runtime dependencies (`mpg123-decoder`, `silk-wasm`, `ws`) that are already mirrored in the root `package.json`. However, qqbot is not present in:
1. `src/plugins/contracts/package-manifest.contract.test.ts` -- validates that plugin `... | closed | completed | true | 4 | [] | [] | 2026-04-13T10:01:47Z | 2026-04-30T04:35:48Z | 2026-04-27T07:58:35Z | CONTRIBUTOR | null | 20260430T060018Z | 2026-04-30T06:00:18Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,405,152 | I_kwDOQb6kR879hc_g | 65,852 | https://github.com/openclaw/openclaw/issues/65852 | https://api.github.com/repos/openclaw/openclaw/issues/65852 | synology-chat missing from .github/labeler.yml | ## Description
`extensions/synology-chat/` and `docs/channels/synology-chat.md` exist, but `.github/labeler.yml` has no `channel: synology-chat` rule. All other bundled channel extensions have corresponding labeler entries.
## Impact
PRs that touch `extensions/synology-chat/**` or `docs/channels/synology-chat.md` do... | closed | completed | true | 2 | [] | [] | 2026-04-13T10:01:55Z | 2026-04-30T04:35:47Z | 2026-04-27T07:58:38Z | CONTRIBUTOR | null | 20260430T060018Z | 2026-04-30T06:00:18Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,406,520 | I_kwDOQb6kR879hdU4 | 65,854 | https://github.com/openclaw/openclaw/issues/65854 | https://api.github.com/repos/openclaw/openclaw/issues/65854 | Video-only providers (runway, alibaba) default to text-inference onboarding scope | ## Description
`extensions/runway/openclaw.plugin.json` and `extensions/alibaba/openclaw.plugin.json` do not set `onboardingScopes` in their `providerAuthChoices`. Per `src/plugins/manifest.ts`, this defaults to `["text-inference"]`.
Both are **video-generation-only** providers (no `providers` key for text, only `con... | closed | completed | true | 1 | [] | [] | 2026-04-13T10:02:07Z | 2026-04-28T04:45:59Z | 2026-04-25T21:53:02Z | CONTRIBUTOR | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,483,890 | I_kwDOQb6kR879hwNy | 65,859 | https://github.com/openclaw/openclaw/issues/65859 | https://api.github.com/repos/openclaw/openclaw/issues/65859 | [Feature]: Add audio upload and playback support in webchat | ### Summary
Add audio file type support for webchat file picker uploads and inline audio playback for received audio attachments.
### Problem to solve
Webchat has two audio-related limitations:
**Upload**: The file picker's accept attribute excludes audio MIME types (e.g., `audio/mpeg`, `audio/mp4`, `audio/ogg`). U... | open | null | false | 1 | [
"enhancement"
] | [] | 2026-04-13T10:12:37Z | 2026-05-02T16:55:32Z | null | NONE | null | 20260502T180029Z | 2026-05-02T18:00:29Z | le-shi | 22,446,162 | MDQ6VXNlcjIyNDQ2MTYy | User | false |
openclaw/openclaw | 4,253,496,726 | I_kwDOQb6kR879hzWW | 65,860 | https://github.com/openclaw/openclaw/issues/65860 | https://api.github.com/repos/openclaw/openclaw/issues/65860 | [Bug]: Error message references nonexistent docs anchor `tools/plugin#runtime-helpers` | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
The error message in `src/gateway/server-plugins.ts` (line 165) references the docs anchor `https://docs.openclaw.ai/tools/plugin#runtime-helpers`, but the target page `docs/tools/plugin.md` does not contain a ... | open | null | false | 2 | [] | [] | 2026-04-13T10:14:05Z | 2026-05-03T04:43:34Z | null | CONTRIBUTOR | null | 20260503T060030Z | 2026-05-03T06:00:30Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,500,229 | I_kwDOQb6kR879h0NF | 65,861 | https://github.com/openclaw/openclaw/issues/65861 | https://api.github.com/repos/openclaw/openclaw/issues/65861 | [Bug]: labeler.yml rule `extensions: minimax-portal-auth` points to nonexistent directory | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
`.github/labeler.yml` (line 284-287) contains a rule for `"extensions: minimax-portal-auth"` that matches `extensions/minimax-portal-auth/**`, but the directory `extensions/minimax-portal-auth/` does not exist ... | closed | not_planned | true | 2 | [] | [] | 2026-04-13T10:14:23Z | 2026-04-30T04:40:57Z | 2026-04-26T05:00:48Z | CONTRIBUTOR | null | 20260430T060018Z | 2026-04-30T06:00:18Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,504,720 | I_kwDOQb6kR879h1TQ | 65,862 | https://github.com/openclaw/openclaw/issues/65862 | https://api.github.com/repos/openclaw/openclaw/issues/65862 | [Bug]: DuckDuckGo web search provider missing from onboarding setup wizard | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
DuckDuckGo web search provider does not appear in the interactive web search setup wizard (`openclaw configure` → web search section) because `createDuckDuckGoWebSearchProvider()` does not set the `onboardingSc... | closed | completed | true | 3 | [] | [] | 2026-04-13T10:14:46Z | 2026-05-04T04:34:05Z | 2026-05-02T03:55:03Z | CONTRIBUTOR | null | 20260504T060031Z | 2026-05-04T06:00:31Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,507,911 | I_kwDOQb6kR879h2FH | 65,863 | https://github.com/openclaw/openclaw/issues/65863 | https://api.github.com/repos/openclaw/openclaw/issues/65863 | [Bug]: docs/providers/index.md missing Bedrock Mantle provider entry | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
`docs/providers/index.md` lists all supported provider doc pages, but it does not include a link to the existing `docs/providers/bedrock-mantle.md` page (Amazon Bedrock Mantle — a distinct provider from the sta... | closed | completed | true | 2 | [] | [] | 2026-04-13T10:15:02Z | 2026-04-28T04:56:37Z | 2026-04-25T17:28:34Z | CONTRIBUTOR | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,511,416 | I_kwDOQb6kR879h274 | 65,864 | https://github.com/openclaw/openclaw/issues/65864 | https://api.github.com/repos/openclaw/openclaw/issues/65864 | [Feature]: Document ZALOUSER_PROFILE and ZCA_PROFILE env vars in Zalo Personal channel docs | ### Summary
Document the `ZALOUSER_PROFILE` and `ZCA_PROFILE` environment variables in `docs/channels/zalouser.md`. These env vars are declared in the plugin manifest and actively read by the code, but the channel docs page does not mention them.
### Problem to solve
Users setting up the Zalo Personal (`zalouser`) c... | closed | not_planned | true | 2 | [] | [] | 2026-04-13T10:15:20Z | 2026-04-30T04:40:57Z | 2026-04-26T05:00:52Z | CONTRIBUTOR | null | 20260430T060018Z | 2026-04-30T06:00:18Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,514,879 | I_kwDOQb6kR879h3x_ | 65,865 | https://github.com/openclaw/openclaw/issues/65865 | https://api.github.com/repos/openclaw/openclaw/issues/65865 | [Feature]: Add feishu_sheets tool support for Feishu Spreadsheet | ### Summary
Add feishu_sheets tool support for Feishu Spreadsheet
### Problem to solve
## Feature Request: feishu_sheets tool
### Description
Add support for reading and writing Feishu Spreadsheets (电子表格) similar to existing feishu_bitable tool.
### Use Case
Users need to:
- Read data from Feishu Spreadsheets
- Wr... | closed | not_planned | true | 1 | [
"enhancement"
] | [] | 2026-04-13T10:15:38Z | 2026-04-28T04:56:37Z | 2026-04-25T17:28:40Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | zhenfengYu | 139,319,657 | U_kgDOCE3ZaQ | User | false |
openclaw/openclaw | 4,253,515,041 | I_kwDOQb6kR879h30h | 65,866 | https://github.com/openclaw/openclaw/issues/65866 | https://api.github.com/repos/openclaw/openclaw/issues/65866 | [Feature]: Document MiniMax TTS environment variables (MINIMAX_TTS_MODEL, MINIMAX_TTS_VOICE_ID) | ### Summary
Document the `MINIMAX_TTS_MODEL` and `MINIMAX_TTS_VOICE_ID` environment variables in `docs/providers/minimax.md`. These are read by the MiniMax speech provider to customize TTS model and voice selection, but neither is documented on the provider page.
### Problem to solve
Users who want to customize Mini... | closed | completed | true | 2 | [] | [] | 2026-04-13T10:15:39Z | 2026-04-27T04:34:51Z | 2026-04-25T03:53:46Z | CONTRIBUTOR | null | 20260427T090020Z | 2026-04-27T09:00:20Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,515,778 | I_kwDOQb6kR879h4AC | 65,867 | https://github.com/openclaw/openclaw/issues/65867 | https://api.github.com/repos/openclaw/openclaw/issues/65867 | [Bug]: Gemini <final> tags leak into delivered messages | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Sorry i am new to the this is a resurfaced issue like #48587 as far as i could gather. everything was fine did newest update now i see the <final> tags in answer of chatbot in webui no tags in whatsapp
### Steps to reprodu... | open | null | false | 18 | [
"bug",
"regression"
] | [] | 2026-04-13T10:15:42Z | 2026-05-02T17:55:05Z | null | NONE | null | 20260502T180029Z | 2026-05-02T18:00:29Z | grizdum | 275,247,635 | U_kgDOEGfyEw | User | false |
openclaw/openclaw | 4,253,519,939 | I_kwDOQb6kR879h5BD | 65,868 | https://github.com/openclaw/openclaw/issues/65868 | https://api.github.com/repos/openclaw/openclaw/issues/65868 | [Feature]: Improve QQBot error messages with actionable guidance and docs links | ### Summary
QQBot extension error messages are overly technical — they dump raw JSON API responses and lack actionable user guidance, documentation links, or env var hints.
### Problem to solve
When QQBot encounters configuration or API errors, the error messages are developer-oriented and unhelpful for end users:
... | open | null | false | 2 | [] | [] | 2026-04-13T10:16:05Z | 2026-05-02T15:22:47Z | null | CONTRIBUTOR | null | 20260502T180029Z | 2026-05-02T18:00:29Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,520,569 | I_kwDOQb6kR879h5K5 | 65,869 | https://github.com/openclaw/openclaw/issues/65869 | https://api.github.com/repos/openclaw/openclaw/issues/65869 | Control UI Dreaming tab calls non-existent wiki.importInsights and wiki.palace APIs | **Description**
Control UI's Dreaming tab attempts to call `wiki.importInsights` and `wiki.palace` methods, but memory-core plugin does not implement these APIs, resulting in `unknown method` errors.
**Steps to Reproduce**
1. Open Control UI (any platform)
2. Navigate to the Dreaming tab
3. Attempt to view dream diary... | closed | completed | true | 3 | [] | [] | 2026-04-13T10:16:09Z | 2026-04-16T04:24:21Z | 2026-04-13T20:02:18Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | kikibrian | 275,287,238 | U_kgDOEGiMxg | User | false |
openclaw/openclaw | 4,253,527,816 | I_kwDOQb6kR879h68I | 65,870 | https://github.com/openclaw/openclaw/issues/65870 | https://api.github.com/repos/openclaw/openclaw/issues/65870 | [Bug]: Brave Search docs URL in code points to legacy path instead of canonical location | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
The Brave web search provider code references `docsUrl: "https://docs.openclaw.ai/brave-search"`, which is the legacy path. While a redirect exists in `docs.json` from `/brave-search` to `/tools/brave-search`, ... | closed | completed | true | 3 | [] | [] | 2026-04-13T10:16:53Z | 2026-05-04T04:34:11Z | 2026-05-02T03:43:28Z | CONTRIBUTOR | null | 20260504T060031Z | 2026-05-04T06:00:31Z | Jah-yee | 166,608,075 | U_kgDOCe48yw | User | false |
openclaw/openclaw | 4,253,592,263 | I_kwDOQb6kR879iKrH | 65,875 | https://github.com/openclaw/openclaw/issues/65875 | https://api.github.com/repos/openclaw/openclaw/issues/65875 | [Bug] Session model override not cleared on /new - falls back to openrouter/free instead of configured primary model | ## Bug Description
**Environment:** OpenClaw 2026.4.11 (Mac macOS)
**Severity:** Medium
## Steps to Reproduce
1. Configure `agents.defaults.model.primary` to a specific model (e.g. \`minimax-portal/MiniMax-M2.7\`) in \`openclaw.json\`
2. Use the assistant normally until a rate limit or model switch occurs, which tri... | closed | completed | true | 1 | [] | [] | 2026-04-13T10:24:06Z | 2026-04-28T04:56:36Z | 2026-04-25T17:28:47Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | marsdream | 19,683,097 | MDQ6VXNlcjE5NjgzMDk3 | User | false |
openclaw/openclaw | 4,253,676,926 | I_kwDOQb6kR879ifV- | 65,883 | https://github.com/openclaw/openclaw/issues/65883 | https://api.github.com/repos/openclaw/openclaw/issues/65883 | webchat: queued user message vanishes from chat history after being dequeued | ## Summary
When the gateway is actively streaming a long assistant response in webchat and the user sends a **second** message, that message is correctly shown as *queued* in the UI. Once the long response finishes and the queued message is dequeued and answered, the assistant's new turn appears — but **the user's que... | closed | completed | true | 1 | [] | [] | 2026-04-13T10:33:21Z | 2026-04-30T04:40:56Z | 2026-04-26T05:00:57Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | faux123 | 598,478 | MDQ6VXNlcjU5ODQ3OA== | User | false |
openclaw/openclaw | 4,253,698,623 | I_kwDOQb6kR879iko_ | 65,887 | https://github.com/openclaw/openclaw/issues/65887 | https://api.github.com/repos/openclaw/openclaw/issues/65887 | [Feature]: Request iOS TestFlight Access | ### Summary
Request access to the official OpenClaw iOS app via TestFlight.
### Problem to solve
I want to use my iPhone as an OpenClaw node, but the iOS app is not publicly distributed yet.
Then scroll down and fill the remaining boxes with this style:
Proposed solution:
Please add me to the iOS TestFlight beta if ... | closed | completed | true | 2 | [
"enhancement",
"r: testflight"
] | [] | 2026-04-13T10:36:11Z | 2026-04-16T04:24:41Z | 2026-04-13T10:36:24Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | GalacticGoat | 42,498,207 | MDQ6VXNlcjQyNDk4MjA3 | User | false |
openclaw/openclaw | 4,253,737,486 | I_kwDOQb6kR879iuIO | 65,888 | https://github.com/openclaw/openclaw/issues/65888 | https://api.github.com/repos/openclaw/openclaw/issues/65888 | [Bug]: gateway-watch-regression check-additional fails across unrelated PRs with dirty_watched_tree invalidation | ## Summary
`check-additional` is failing across multiple unrelated PRs because `gateway-watch-regression` invalidates early with `dirty_watched_tree`, even when the PR diff does not touch gateway watch logic and local reproduction on the PR head passes.
## Evidence
Observed on these unrelated PRs:
- `#65771` fix(cron)... | closed | completed | true | 1 | [] | [] | 2026-04-13T10:40:54Z | 2026-04-28T04:56:35Z | 2026-04-25T17:28:53Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | yozu | 1,403,979 | MDQ6VXNlcjE0MDM5Nzk= | User | false |
openclaw/openclaw | 4,253,751,294 | I_kwDOQb6kR879ixf- | 65,890 | https://github.com/openclaw/openclaw/issues/65890 | https://api.github.com/repos/openclaw/openclaw/issues/65890 | [Bug]: QMD backend falls back to builtin on live `openclaw memory search` due to managed collection mismatch (`memory-alt-main` vs `memory-root-main`) | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
With `memory.backend = "qmd"`, live `openclaw memory search` falls back to builtin because OpenClaw requests `memory-alt-main` while the managed QMD state reports an existing collection named `memory-root-main` for the same ... | closed | completed | true | 2 | [
"bug",
"regression"
] | [] | 2026-04-13T10:42:37Z | 2026-04-16T04:23:21Z | 2026-04-13T22:00:45Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | chrissuarez | 10,571,531 | MDQ6VXNlcjEwNTcxNTMx | User | false |
openclaw/openclaw | 4,253,859,097 | I_kwDOQb6kR879jL0Z | 65,897 | https://github.com/openclaw/openclaw/issues/65897 | https://api.github.com/repos/openclaw/openclaw/issues/65897 | npm audit: 12 unresolved vulnerabilities in deep transitive dependencies | ## Problem
After `npm install openclaw@latest` and `npm audit fix --force`, 12 vulnerabilities remain in deep transitive dependencies that cannot be fixed without breaking the parent packages:
### Critical (3)
- **axios <1.15.0** via @larksuiteoapi/node-sdk — SSRF/NOH-PROXY bypass (GHSA-3p68-rc4w-qgx5, GHSA-fvcv-3m26... | closed | completed | true | 1 | [] | [] | 2026-04-13T10:55:51Z | 2026-04-23T04:22:34Z | 2026-04-20T14:37:37Z | NONE | null | 20260423T060020Z | 2026-04-23T06:00:20Z | lyh-github-97 | 73,977,041 | MDQ6VXNlcjczOTc3MDQx | User | false |
openclaw/openclaw | 4,253,873,330 | I_kwDOQb6kR879jPSy | 65,899 | https://github.com/openclaw/openclaw/issues/65899 | https://api.github.com/repos/openclaw/openclaw/issues/65899 | Hook error: Cannot access 'group' before initialization in message:sent | ## Description
Internal hook `message:sent` throws `ReferenceError: Cannot access 'group' before initialization` on every message send. The error occurs in a managed/built-in hook (not user code).
## Version
OpenClaw `2026.4.11` (769908e)
## Steps to Reproduce
1. Send any message through the gateway (e.g., via Tel... | closed | completed | true | 1 | [] | [] | 2026-04-13T10:57:28Z | 2026-04-28T04:56:34Z | 2026-04-25T17:29:07Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | artdaal | 35,874,009 | MDQ6VXNlcjM1ODc0MDA5 | User | false |
openclaw/openclaw | 4,253,966,963 | I_kwDOQb6kR879jmJz | 65,902 | https://github.com/openclaw/openclaw/issues/65902 | https://api.github.com/repos/openclaw/openclaw/issues/65902 | [Bug]: WhatsApp session discovery inconsistent with runtime status and cron delivery | ## Summary
On OpenClaw 2026.4.11 on macOS, WhatsApp delivery via cron works, and `openclaw status --json` shows a direct WhatsApp session (`agent:main:whatsapp:direct:+5511996700808`). However, agent-side session discovery for on-demand automation appears inconsistent: earlier automated flows and `sessions_list`-equiv... | open | null | false | 2 | [] | [] | 2026-04-13T11:07:57Z | 2026-05-02T19:50:19Z | null | NONE | null | 20260502T210030Z | 2026-05-02T21:00:30Z | vinivazquez-dealerspace | 269,463,143 | U_kgDOEA-uZw | User | false |
openclaw/openclaw | 4,253,994,008 | I_kwDOQb6kR879jswY | 65,903 | https://github.com/openclaw/openclaw/issues/65903 | https://api.github.com/repos/openclaw/openclaw/issues/65903 | Docs: Create runbooks folder structure for all tools and systems | ## Context
We need a runbooks folder in the docs with per-tool/per-system runbooks so we can:
- Document how to use each tool (MiniMax TTS, image, video, music; OpenClaw; Trading Bot)
- Log issues we hit and how we fixed them
- Document configuration steps so the knowledge persists
## Proposed Structure
`docs/runbooks... | closed | completed | true | 1 | [] | [] | 2026-04-13T11:10:30Z | 2026-04-28T04:49:41Z | 2026-04-25T19:04:24Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | SimulatedPixels | 257,303,797 | U_kgDOD1Yk9Q | User | false |
openclaw/openclaw | 4,254,059,023 | I_kwDOQb6kR879j8oP | 65,905 | https://github.com/openclaw/openclaw/issues/65905 | https://api.github.com/repos/openclaw/openclaw/issues/65905 | [Bug] Windows npm global install: extensionAPI.js cannot be resolved by plugins | ## Bug Description
**Environment:**
- OS: Windows (npm global install)
- OpenClaw: 2026.4.9
- Plugin: memory-lancedb-pro v1.1.0-beta.10
**Problem:**
On Windows with npm global install, the `extensionAPI.js` file cannot be resolved by plugins that try to dynamically import it.
**Error Message:**
```
fallback used for... | closed | completed | true | 1 | [] | [] | 2026-04-13T11:17:20Z | 2026-04-28T04:56:33Z | 2026-04-25T17:29:14Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | jlin53882 | 76,562,059 | MDQ6VXNlcjc2NTYyMDU5 | User | false |
openclaw/openclaw | 4,254,100,680 | I_kwDOQb6kR879kGzI | 65,908 | https://github.com/openclaw/openclaw/issues/65908 | https://api.github.com/repos/openclaw/openclaw/issues/65908 | memory search still targets stale QMD collections (memory-alt-main/custom-1-main) and falls back to builtin | ## Summary
`openclaw memory search` can still target stale/missing managed QMD collection names even when the live QMD config and index are healthy. When this happens, memory recall still works via builtin fallback, but the command emits noisy repair/failure output and can trip failure-only canaries.
## Version
- Open... | closed | completed | true | 2 | [] | [] | 2026-04-13T11:24:12Z | 2026-04-28T04:56:32Z | 2026-04-25T17:29:20Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | KZMSolutions | 45,105,356 | MDQ6VXNlcjQ1MTA1MzU2 | User | false |
openclaw/openclaw | 4,254,105,422 | I_kwDOQb6kR879kH9O | 65,909 | https://github.com/openclaw/openclaw/issues/65909 | https://api.github.com/repos/openclaw/openclaw/issues/65909 | Codex native web_search does not reliably activate despite valid config/auth | ## Summary
With Codex native `web_search` enabled in OpenClaw, native Codex search does not appear to activate reliably in practice, even though the config and activation conditions look correct from the code.
I am opening this to clarify expected behavior and whether there is a bug or an additional condition not doc... | closed | completed | true | 2 | [] | [] | 2026-04-13T11:24:57Z | 2026-05-03T04:35:25Z | 2026-04-30T19:18:10Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | ismael-81 | 273,866,131 | U_kgDOEFLdkw | User | false |
openclaw/openclaw | 4,254,120,869 | I_kwDOQb6kR879kLul | 65,910 | https://github.com/openclaw/openclaw/issues/65910 | https://api.github.com/repos/openclaw/openclaw/issues/65910 | Memory recall: read/exec tool usage does not generate recall signals | ## Problem
The short-term-recall.json (memory/.dreams/) only records recall signals when the agent explicitly calls the memory search tool. Reading source files, running exec commands, or querying DBs — even when they verify or discover durable facts — do not generate any recall signal.
**Current flow:**
1. Agent rea... | closed | not_planned | true | 0 | [] | [] | 2026-04-13T11:27:46Z | 2026-04-16T04:24:40Z | 2026-04-13T11:34:16Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | patng-max | 228,122,423 | U_kgDODZjfNw | User | false |
openclaw/openclaw | 4,254,126,009 | I_kwDOQb6kR879kM-5 | 65,912 | https://github.com/openclaw/openclaw/issues/65912 | https://api.github.com/repos/openclaw/openclaw/issues/65912 | Slack provider: thread_ts corrupted to Int64.MAX_VALUE (9223372036854775807) on threaded reply — invalid_thread_ts API error | **Version:** 2026.4.11 (769908e)
**Platform:** macOS 26.3.1 arm64 · Node 25.9.0
## What happened
Slack DM messages stopped receiving responses intermittently. The gateway
remained running. Gateway logs contained:
```
ERROR: 9223372036854775807 is not a valid thread_ts
Error: An API error occurred: invalid_thread_ts
... | closed | completed | true | 2 | [] | [] | 2026-04-13T11:28:42Z | 2026-04-28T04:56:32Z | 2026-04-25T17:29:34Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | redasadki | 5,021,946 | MDQ6VXNlcjUwMjE5NDY= | User | false |
openclaw/openclaw | 4,254,163,910 | I_kwDOQb6kR879kWPG | 65,913 | https://github.com/openclaw/openclaw/issues/65913 | https://api.github.com/repos/openclaw/openclaw/issues/65913 | QMD default memory collection can generate phantom `memory-alt-*` lookups when only `MEMORY.md` exists | ## Summary
With `memory.backend = "qmd"` and `memory.qmd.includeDefaultMemory = true`, OpenClaw can generate and later search a managed collection for lowercase `memory.md` even when the workspace only has uppercase `MEMORY.md`.
That can produce a phantom managed collection lookup like:
```text
Collection not found:... | closed | completed | true | 1 | [] | [] | 2026-04-13T11:35:25Z | 2026-04-16T04:24:18Z | 2026-04-13T22:00:45Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | shawnduggan | 3,504,807 | MDQ6VXNlcjM1MDQ4MDc= | User | false |
openclaw/openclaw | 4,254,196,494 | I_kwDOQb6kR879keMO | 65,915 | https://github.com/openclaw/openclaw/issues/65915 | https://api.github.com/repos/openclaw/openclaw/issues/65915 | [Bug]: [opik-openclaw] Plugin fails to export traces after upgrading from 2026.3.13 to 2026.4.11 (no outgoing requests to Opik) | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
After upgrading OpenClaw from version 2026.3.13 to 2026.4.11, the @opik/opik-openclaw plugin stopped exporting traces to Opik. The plugin status shows as loaded and no errors appear in the logs, but tcpdump confirms that no ... | closed | not_planned | true | 1 | [
"bug",
"regression"
] | [] | 2026-04-13T11:40:49Z | 2026-04-28T04:56:31Z | 2026-04-25T17:29:41Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | zhangyongjun10 | 125,647,109 | U_kgDOB305BQ | User | false |
openclaw/openclaw | 4,254,208,406 | I_kwDOQb6kR879khGW | 65,916 | https://github.com/openclaw/openclaw/issues/65916 | https://api.github.com/repos/openclaw/openclaw/issues/65916 | Cron engine crashes on startup when job has no state field | ## Description
`cron.start()` crashes with `TypeError: Cannot read properties of undefined (reading 'runningAtMs')` when any job in `jobs.json` has a `null` or missing `state` field.
This prevents **all** cron jobs from running — not just the one with the missing state.
## Version
OpenClaw 2026.4.1 (da64a97)
## St... | closed | completed | true | 2 | [] | [] | 2026-04-13T11:42:37Z | 2026-04-28T05:28:45Z | 2026-04-25T09:39:08Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | guillaumeEmery | 23,261,147 | MDQ6VXNlcjIzMjYxMTQ3 | User | false |
openclaw/openclaw | 4,254,217,111 | I_kwDOQb6kR879kjOX | 65,917 | https://github.com/openclaw/openclaw/issues/65917 | https://api.github.com/repos/openclaw/openclaw/issues/65917 | [Bug]: | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
After upgrading from 2026.4.2 to 2026.4.11, QQ Bot fails to download images with [security] blocked URL fetch (url-fetch) error; files download correctly but images are blocked by SSRF guard.
### Steps to reproduce
1、Start... | closed | completed | true | 2 | [
"bug",
"regression"
] | [] | 2026-04-13T11:43:58Z | 2026-04-17T04:22:18Z | 2026-04-15T04:07:04Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | x0620x | 54,536,082 | MDQ6VXNlcjU0NTM2MDgy | User | false |
openclaw/openclaw | 4,254,230,454 | I_kwDOQb6kR879kme2 | 65,918 | https://github.com/openclaw/openclaw/issues/65918 | https://api.github.com/repos/openclaw/openclaw/issues/65918 | openclaw config: OpenAI provider wizard shows misleading hint "Codex OAuth + API key" | ## Bug Description
When running `openclaw config` to set up a model provider, the interactive wizard presents **OpenAI** and **OpenAI Codex** as two separate provider options. However, the **OpenAI** option displays the group hint as **"Codex OAuth + API key"**, which is misleading because:
1. OpenAI and OpenAI Codex... | closed | completed | true | 1 | [] | [] | 2026-04-13T11:46:09Z | 2026-04-27T04:41:07Z | 2026-04-25T02:12:45Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | jojojonathan | 37,664,587 | MDQ6VXNlcjM3NjY0NTg3 | User | false |
openclaw/openclaw | 4,254,265,604 | I_kwDOQb6kR879kvEE | 65,919 | https://github.com/openclaw/openclaw/issues/65919 | https://api.github.com/repos/openclaw/openclaw/issues/65919 | [Feature]: Keep configured presence alive until run complete + dispatch idle | ### Summary
Keep configured active-run presence such as typing, reactions, or existing channel progress surfaces alive until the run is actually complete and reply dispatch is idle.
### Problem to solve
Healthy long turns can lose visible presence before they actually finish.
Concrete examples:
- Telegram: a user ... | closed | completed | true | 4 | [
"enhancement"
] | [] | 2026-04-13T11:49:44Z | 2026-05-05T04:25:46Z | 2026-05-03T02:04:05Z | CONTRIBUTOR | null | 20260505T060034Z | 2026-05-05T06:00:34Z | rubencu | 4,742,789 | MDQ6VXNlcjQ3NDI3ODk= | User | false |
openclaw/openclaw | 4,254,271,525 | I_kwDOQb6kR879kwgl | 65,920 | https://github.com/openclaw/openclaw/issues/65920 | https://api.github.com/repos/openclaw/openclaw/issues/65920 | openclaw config overwrites existing models list instead of merging | ## Bug Description
When using `openclaw config` to add a new provider and select models, the `agents.defaults.models` (or equivalent) array in `openclaw.json` gets **completely overwritten** with only the newly selected models. All previously configured model entries are lost.
## Steps to Reproduce
1. Start with an ... | closed | completed | true | 1 | [] | [] | 2026-04-13T11:50:22Z | 2026-04-25T04:16:22Z | 2026-04-22T22:18:23Z | NONE | null | 20260425T060021Z | 2026-04-25T06:00:21Z | jojojonathan | 37,664,587 | MDQ6VXNlcjM3NjY0NTg3 | User | false |
openclaw/openclaw | 4,254,311,541 | I_kwDOQb6kR879k6R1 | 65,921 | https://github.com/openclaw/openclaw/issues/65921 | https://api.github.com/repos/openclaw/openclaw/issues/65921 | [Bug]: ACP launcher resolves bundled acpx from wrong path after update | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Version: OpenClaw 2026.4.11 (769908e)
ACP launch fails because OpenClaw resolves the bundled `acpx` binary from a non-existent nested plugin path.
Exact failing path:
`/home/anon/.nvm/versions/node/v25.6.1/lib/node_modules/... | closed | completed | true | 2 | [
"bug"
] | [] | 2026-04-13T11:54:29Z | 2026-04-28T04:56:30Z | 2026-04-25T17:29:47Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | kvzsolt | 32,572,393 | MDQ6VXNlcjMyNTcyMzkz | User | false |
openclaw/openclaw | 4,254,376,249 | I_kwDOQb6kR879lKE5 | 65,923 | https://github.com/openclaw/openclaw/issues/65923 | https://api.github.com/repos/openclaw/openclaw/issues/65923 | session.reset unset still defaults to silent daily reset at 4am | ## Bug / UX footgun
When `session.reset` is not configured at all, OpenClaw still applies a default reset policy of:
- `mode: \"daily\"`
- `atHour: 4`
This is highly surprising and easy to miss. Most operators expect **unset** to mean **no automatic reset**.
## Why this is a problem
This leads to silent session ro... | closed | not_planned | true | 1 | [] | [] | 2026-04-13T12:01:32Z | 2026-05-03T04:35:32Z | 2026-04-30T18:45:09Z | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | jojojonathan | 37,664,587 | MDQ6VXNlcjM3NjY0NTg3 | User | false |
openclaw/openclaw | 4,254,456,794 | I_kwDOQb6kR879ldva | 65,924 | https://github.com/openclaw/openclaw/issues/65924 | https://api.github.com/repos/openclaw/openclaw/issues/65924 | Retry on engine_overloaded stores both failed and successful assistant messages, creating duplicate tool_call IDs | ## Bug Description
When the LLM provider returns a transient error (e.g., `engine_overloaded`), OpenClaw retries the request. However, **both the failed response and the successful retry are persisted to the session JSONL**, and if the failed response contained `toolCall` entries, the retry often produces the same too... | closed | completed | true | 1 | [] | [] | 2026-04-13T12:09:54Z | 2026-04-30T04:40:55Z | 2026-04-26T05:01:01Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | buxibuxi | 3,839,922 | MDQ6VXNlcjM4Mzk5MjI= | User | false |
openclaw/openclaw | 4,254,488,496 | I_kwDOQb6kR879llew | 65,934 | https://github.com/openclaw/openclaw/issues/65934 | https://api.github.com/repos/openclaw/openclaw/issues/65934 | Memory dreaming: recall signals only from memory search, not from read/exec | ## Problem
The built-in dreaming promotion scores candidate facts from daily memory notes based on recall signals stored in `short-term-recall.json`. However, these recall signals are **only generated when the memory search tool is explicitly called**.
When an agent discovers or verifies a fact by reading a source fi... | closed | duplicate | true | 0 | [] | [] | 2026-04-13T12:13:23Z | 2026-04-17T04:22:53Z | 2026-04-14T11:29:37Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | patng-max | 228,122,423 | U_kgDODZjfNw | User | false |
openclaw/openclaw | 4,254,506,099 | I_kwDOQb6kR879lpxz | 65,935 | https://github.com/openclaw/openclaw/issues/65935 | https://api.github.com/repos/openclaw/openclaw/issues/65935 | fix: keep trade engine online across runtime restarts | ## Problem Statement
The owner overlay had a long-lived , but no repo-managed wrapper. After a manual stop or a reboot, the bridge could stay up while the custom trade engine remained down, so and bridge health reported the engine as unavailable until someone started it manually.
## Proposal
Add a user-systemd , a... | closed | completed | true | 1 | [] | [] | 2026-04-13T12:15:33Z | 2026-04-16T04:24:40Z | 2026-04-13T12:17:26Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | phongquach239 | 197,099,961 | U_kgDOC7-BuQ | User | false |
openclaw/openclaw | 4,254,564,828 | I_kwDOQb6kR879l4Hc | 65,945 | https://github.com/openclaw/openclaw/issues/65945 | https://api.github.com/repos/openclaw/openclaw/issues/65945 | [Feature]: Auto‑forward images from Node agent.request to IM users when deliver: true | ### Summary
When a Node node pushes an image to the Gateway via the agent.request event (with deliver: true), the image is correctly received and processed by the Gateway (inline/offload) and can be analysed by the Agent. However, the image is not automatically sent to the target IM user (e.g. QQ). The Agent must expl... | open | null | false | 1 | [
"enhancement"
] | [] | 2026-04-13T12:22:30Z | 2026-05-03T01:22:41Z | null | NONE | null | 20260503T030030Z | 2026-05-03T03:00:30Z | Yanxingang | 17,355,243 | MDQ6VXNlcjE3MzU1MjQz | User | false |
openclaw/openclaw | 4,254,602,105 | I_kwDOQb6kR879mBN5 | 65,946 | https://github.com/openclaw/openclaw/issues/65946 | https://api.github.com/repos/openclaw/openclaw/issues/65946 | Skills from extraDirs and workspace symlinks register as eligible but are not injected into agent sessions | ## Description
Skills loaded via `skills.load.extraDirs` (and workspace `skills/` symlinks) appear as `eligible: true`, pass `skills check`, and show as `ready` in `skills list`, but are **never injected into agent session prompts**. The skill's `SKILL.md` content does not appear in the agent's system prompt `skills.e... | closed | completed | true | 1 | [] | [] | 2026-04-13T12:27:21Z | 2026-04-28T04:56:29Z | 2026-04-25T17:29:54Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | stamkivi | 5,263,291 | MDQ6VXNlcjUyNjMyOTE= | User | false |
openclaw/openclaw | 4,254,682,966 | I_kwDOQb6kR879mU9W | 65,948 | https://github.com/openclaw/openclaw/issues/65948 | https://api.github.com/repos/openclaw/openclaw/issues/65948 | Discord file send via message tool fails: 'empty message' error even with file attached |
## Bug Description
When using the `message` tool's `filePath` or `media` parameter to send a file to Discord, it fails with `Cannot send an empty message` even when:
1. The file path is inside the allowed workspace directory (`~/.openclaw/workspace/`)
2. A non-empty text message is provided alongside the file
## Ste... | closed | completed | true | 1 | [] | [] | 2026-04-13T12:38:53Z | 2026-04-28T04:56:29Z | 2026-04-25T17:30:00Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | TBodyAltra | 125,636,483 | U_kgDOB30Pgw | User | false |
openclaw/openclaw | 4,254,727,434 | I_kwDOQb6kR879mf0K | 65,949 | https://github.com/openclaw/openclaw/issues/65949 | https://api.github.com/repos/openclaw/openclaw/issues/65949 | [Bug]: Existing Discord thread sessions do not pick up newly working browser capability and continue reporting no browser tool present | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
In Discord threads with prior workflow history, OpenClaw continues reporting no browser tool present even after browser automation is configured and verified working on the host.
### Steps to reproduce
1. Ru... | closed | completed | true | 2 | [
"bug",
"bug:behavior"
] | [] | 2026-04-13T12:45:17Z | 2026-04-28T04:56:28Z | 2026-04-25T17:30:07Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Justin2172021 | 95,263,233 | U_kgDOBa2aAQ | User | false |
openclaw/openclaw | 4,254,780,274 | I_kwDOQb6kR879msty | 65,950 | https://github.com/openclaw/openclaw/issues/65950 | https://api.github.com/repos/openclaw/openclaw/issues/65950 | [Bug]: Isolated cron sessions falsely report status:ok despite incomplete work | ### Bug type
Regression (worked before, now fails)
### Beta release blocker
No
### Summary
Isolated cron sessions report `status: ok` and `delivered: true` in the cron run history even when the agent never completes its intended workflow.
### Steps to reproduce
1. Create a cron job with `sessionTarget: "isolated... | closed | completed | true | 2 | [
"bug",
"regression"
] | [] | 2026-04-13T12:54:57Z | 2026-04-27T04:43:20Z | 2026-04-25T01:11:17Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | hexsigil | 121,487,962 | U_kgDOBz3CWg | User | false |
openclaw/openclaw | 4,254,849,214 | I_kwDOQb6kR879m9i- | 65,951 | https://github.com/openclaw/openclaw/issues/65951 | https://api.github.com/repos/openclaw/openclaw/issues/65951 | [Bug]: Telegram DM voice-note replies skip auto TTS on 2026.4.11 when transcript replaces <media:audio> and inboundAudio becomes false | ## Summary
On OpenClaw `2026.4.11`, Telegram direct-chat voice notes are transcribed successfully, but auto voice reply with:
```json
{"messages":{"tts":{"auto":"inbound"}}}
```
does not trigger.
Observed behaviour:
- inbound Telegram voice note arrives
- STT / preflight transcription works
- assistant replies in p... | closed | completed | true | 1 | [] | [] | 2026-04-13T13:05:05Z | 2026-04-30T04:45:19Z | 2026-04-25T06:46:17Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | GeoKM | 45,901 | MDQ6VXNlcjQ1OTAx | User | false |
openclaw/openclaw | 4,254,985,713 | I_kwDOQb6kR879ne3x | 65,953 | https://github.com/openclaw/openclaw/issues/65953 | https://api.github.com/repos/openclaw/openclaw/issues/65953 | test | test | closed | not_planned | true | 1 | [] | [] | 2026-04-13T13:17:58Z | 2026-04-28T04:56:27Z | 2026-04-25T17:30:14Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | donsummerwind | 6,355,580 | MDQ6VXNlcjYzNTU1ODA= | User | false |
openclaw/openclaw | 4,254,987,294 | I_kwDOQb6kR879nfQe | 65,954 | https://github.com/openclaw/openclaw/issues/65954 | https://api.github.com/repos/openclaw/openclaw/issues/65954 | [Bug]: visible completion, transcript persistence, and delivery seem broken across the same async/isolated seam | ## Summary
There appears to be a broader OpenClaw bug cluster around **successful async/CLI/isolated completions being visible to the user but not reliably available to later system paths**.
This is showing up across multiple surfaces that currently look separate in issue tracker terms:
- follow-up isolated agent run... | closed | completed | true | 0 | [] | [] | 2026-04-13T13:18:08Z | 2026-04-20T04:26:32Z | 2026-04-17T17:00:30Z | NONE | null | 20260420T060038Z | 2026-04-20T06:00:38Z | aaajiao | 49,885 | MDQ6VXNlcjQ5ODg1 | User | false |
openclaw/openclaw | 4,255,058,033 | I_kwDOQb6kR879nwhx | 65,956 | https://github.com/openclaw/openclaw/issues/65956 | https://api.github.com/repos/openclaw/openclaw/issues/65956 | Session messages not persisted across disconnects — WeCom channel data loss | ## Bug Report: Session messages not persisted across disconnects — WeCom channel data loss
### Description
Session message persistence is broken for the WeCom channel. Messages are **not reliably saved** to session JSONL files when the session disconnects and reconnects. This results in **complete loss of conversatio... | closed | completed | true | 0 | [] | [] | 2026-04-13T13:24:45Z | 2026-04-16T04:24:39Z | 2026-04-13T13:35:22Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | donsummerwind | 6,355,580 | MDQ6VXNlcjYzNTU1ODA= | User | false |
openclaw/openclaw | 4,255,086,311 | I_kwDOQb6kR879n3bn | 65,958 | https://github.com/openclaw/openclaw/issues/65958 | https://api.github.com/repos/openclaw/openclaw/issues/65958 | [Bug]: GitHub skill requires OpenClaw restart after CLI installation but restart guidance is unclear | ### Bug type
Documentation/UX issue
### Summary
After installing GitHub CLI and completing authentication, the GitHub skill functionality may not be fully available until OpenClaw is restarted. The current documentation mentions restart is needed, but:
1. The restart requirement is not prominently highlighted
2. No... | open | null | false | 1 | [] | [] | 2026-04-13T13:27:34Z | 2026-05-02T12:53:55Z | null | NONE | null | 20260502T150030Z | 2026-05-02T15:00:30Z | 416524479-max | 274,206,385 | U_kgDOEFgOsQ | User | false |
openclaw/openclaw | 4,255,097,470 | I_kwDOQb6kR879n6J- | 65,959 | https://github.com/openclaw/openclaw/issues/65959 | https://api.github.com/repos/openclaw/openclaw/issues/65959 | Monetization: Add support for Google AdSense, Apoia.se, donation buttons, Amazon & MercadoLibre affiliate banners | ## Problem Statement
OpenClaw is an open-source project that takes significant effort to maintain. While the project thrives on community contributions, having sustainable funding mechanisms is critical for long-term viability. Many similar open-source projects (Node.js, TensorFlow, Linux Foundation projects) rely on ... | closed | completed | true | 0 | [] | [] | 2026-04-13T13:28:28Z | 2026-04-16T04:24:39Z | 2026-04-13T13:32:38Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | victorborsoi | 2,334,623 | MDQ6VXNlcjIzMzQ2MjM= | User | false |
openclaw/openclaw | 4,255,159,245 | I_kwDOQb6kR879oJPN | 65,961 | https://github.com/openclaw/openclaw/issues/65961 | https://api.github.com/repos/openclaw/openclaw/issues/65961 | [Analysis]: Project health analysis reveals version sync, skill path warnings, and documentation sync issues | ### Analysis Type
Project health and user experience analysis
### Summary
Based on a comprehensive GitHub analysis of the OpenClaw project, I've identified several areas for improvement that affect user experience and project health:
1. **Version synchronization gap**: Users may not be aware of new releases
2. **Sk... | closed | completed | true | 1 | [] | [] | 2026-04-13T13:34:05Z | 2026-04-28T04:56:26Z | 2026-04-25T17:30:20Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | 416524479-max | 274,206,385 | U_kgDOEFgOsQ | User | false |
openclaw/openclaw | 4,255,215,946 | I_kwDOQb6kR879oXFK | 65,963 | https://github.com/openclaw/openclaw/issues/65963 | https://api.github.com/repos/openclaw/openclaw/issues/65963 | Dreaming: session sprawl, missing model override, no auto-cleanup | ## Problem
Dreaming creates a **pair of sessions (narrative-light + narrative-rem) per memory chunk per cycle**. With the default `0 */4 * * *` frequency, this produces **~90+ session pairs per day**, all using the agent's default model (in our case, `claude-opus-4-6`).
These completed dreaming sessions:
- **Clog the... | open | null | false | 2 | [] | [] | 2026-04-13T13:40:20Z | 2026-05-03T03:22:49Z | null | NONE | null | 20260503T060030Z | 2026-05-03T06:00:30Z | esqandil | 243,999,297 | U_kgDODosiQQ | User | false |
openclaw/openclaw | 4,255,218,014 | I_kwDOQb6kR879oXle | 65,964 | https://github.com/openclaw/openclaw/issues/65964 | https://api.github.com/repos/openclaw/openclaw/issues/65964 | TTS tool audio silently dropped: isAllowedAbsoluteReplyMediaPath rejects /tmp/openclaw paths | ## Bug
The built-in `tts` tool generates audio successfully but voice notes are silently dropped during delivery to Matrix (and likely all channels). No error is logged.
## Root Cause
`isAllowedAbsoluteReplyMediaPath()` in `agent-runner.runtime` only allows media from workspace/sandbox dirs and managed global media ... | closed | completed | true | 4 | [] | [] | 2026-04-13T13:40:35Z | 2026-04-27T04:37:58Z | 2026-04-25T03:03:53Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | vkrmch | 6,424,784 | MDQ6VXNlcjY0MjQ3ODQ= | User | false |
openclaw/openclaw | 4,255,249,478 | I_kwDOQb6kR879ofRG | 65,965 | https://github.com/openclaw/openclaw/issues/65965 | https://api.github.com/repos/openclaw/openclaw/issues/65965 | memory-wiki: wiki ingest source slugs silently overwrite CJK titles (ASCII-only slug + no unique suffix) | ## Summary
`openclaw wiki ingest` can silently overwrite existing `Wiki/sources/*.md` pages when source titles contain CJK or other non-ASCII characters, because source identity is derived from an ASCII-only slug.
## Current implementation
In installed 2026.4.11 (`dist/cli-fw2V8Mig.js`):
```js
function slugifyWikiSeg... | closed | completed | true | 1 | [] | [] | 2026-04-13T13:44:45Z | 2026-04-22T04:30:59Z | 2026-04-19T12:58:42Z | NONE | null | 20260422T060039Z | 2026-04-22T06:00:39Z | richardmqq | 14,178,267 | MDQ6VXNlcjE0MTc4MjY3 | User | false |
openclaw/openclaw | 4,255,271,167 | I_kwDOQb6kR879okj_ | 65,967 | https://github.com/openclaw/openclaw/issues/65967 | https://api.github.com/repos/openclaw/openclaw/issues/65967 | TUI model name not updated after fallback - still shows old model instead of actual running model | ## Bug Description
**Phenomenon**: When a model fallback is triggered, the TUI (and session_status) still displays the old default model name instead of the actual fallback model that is running.
**Steps to reproduce**:
1. Configure default model as `ollama/gemma4:e4b`, fallback as `minimax-portal/MiniMax-M2.7-highsp... | closed | completed | true | 1 | [] | [] | 2026-04-13T13:47:09Z | 2026-04-28T04:56:26Z | 2026-04-25T17:30:27Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | zamanewby | 97,544 | MDQ6VXNlcjk3NTQ0 | User | false |
openclaw/openclaw | 4,255,387,730 | I_kwDOQb6kR879pBBS | 65,970 | https://github.com/openclaw/openclaw/issues/65970 | https://api.github.com/repos/openclaw/openclaw/issues/65970 | Active Memory sub-agent returns NONE despite memory_search returning valid results | ## Environment
- OpenClaw v2026.4.11
- Ubuntu 24.04, VM on Proxmox
- Active Memory plugin enabled
- Custom memory backend: AngieMem (sqlite-vec) registered via mcporter as QMD server
- `memory.backend: "qmd"`, `memory.qmd.mcporter.enabled: true`, `memory.qmd.searchTool: "search"`, `memory.qmd.searchMode: "search"`
- `c... | closed | completed | true | 2 | [] | [] | 2026-04-13T14:00:05Z | 2026-04-16T04:24:33Z | 2026-04-13T15:16:41Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | dertbv | 3,845,627 | MDQ6VXNlcjM4NDU2Mjc= | User | false |
openclaw/openclaw | 4,255,395,977 | I_kwDOQb6kR879pDCJ | 65,971 | https://github.com/openclaw/openclaw/issues/65971 | https://api.github.com/repos/openclaw/openclaw/issues/65971 | [Bug]:[macOS] Gateway stops processing messages after 15-20 minutes, requires restart | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Gateway on macOS enters zombie state after 15-20 minutes of operation: process remains alive and port listens, but stops processing all incoming messages with no error logs.
### Steps to reproduce
1. Install... | closed | completed | true | 1 | [
"bug",
"bug:behavior"
] | [] | 2026-04-13T14:01:17Z | 2026-04-28T04:56:25Z | 2026-04-25T17:30:40Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | e1026 | 264,750,316 | U_kgDOD8fE7A | User | false |
openclaw/openclaw | 4,255,454,409 | I_kwDOQb6kR879pRTJ | 65,974 | https://github.com/openclaw/openclaw/issues/65974 | https://api.github.com/repos/openclaw/openclaw/issues/65974 | [Bug]: | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Windows native install on OpenClaw 2026.4.11 is unusable for me.
What happened:
- WSL2 path blocked because ThinkPad BIOS virtualization settings are locked
- Native Windows install reports missing bundled plu... | open | null | false | 1 | [
"bug"
] | [] | 2026-04-13T14:07:35Z | 2026-05-03T02:15:01Z | null | NONE | null | 20260503T030030Z | 2026-05-03T03:00:30Z | faewnkj2 | 210,847,362 | U_kgDODJFGgg | User | false |
openclaw/openclaw | 4,255,512,366 | I_kwDOQb6kR879pfcu | 65,975 | https://github.com/openclaw/openclaw/issues/65975 | https://api.github.com/repos/openclaw/openclaw/issues/65975 | Gateway HTTP endpoints return 502 after upgrading to 2026.4.12 | ## Bug Description
After upgrading from OpenClaw 2026.3.13 to 2026.4.12, all Gateway HTTP endpoints return `502 Bad Gateway`. WebSocket connections (used by QQ Bot and other channels) work fine, but any HTTP request to `http://127.0.0.1:18789/` or `/tools/invoke` fails with 502.
## Steps to Reproduce
1. Have OpenCla... | closed | completed | true | 1 | [] | [] | 2026-04-13T14:14:11Z | 2026-04-28T04:56:24Z | 2026-04-25T17:30:53Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | JasonKong1990 | 18,439,086 | MDQ6VXNlcjE4NDM5MDg2 | User | false |
openclaw/openclaw | 4,255,519,712 | I_kwDOQb6kR879phPg | 65,976 | https://github.com/openclaw/openclaw/issues/65976 | https://api.github.com/repos/openclaw/openclaw/issues/65976 | memory-wiki bridge mode reports zero public artifacts and can remove existing bridge pages even when QMD memory is healthy | ## Title
`memory-wiki` bridge mode reports zero public artifacts and can remove existing bridge pages even when QMD memory is healthy
## Summary
On a local OpenClaw host running `2026.4.12`, `memory-wiki` bridge mode still reports `bridgePublicArtifactCount: 0` through the stock `openclaw wiki` command path, even th... | closed | completed | true | 4 | [
"dedupe:child"
] | [
"vincentkoc"
] | 2026-04-13T14:15:06Z | 2026-04-29T04:29:25Z | 2026-04-26T08:21:36Z | NONE | null | 20260429T060025Z | 2026-04-29T06:00:25Z | carl-jeffrolc | 258,705,842 | U_kgDOD2uJsg | User | false |
openclaw/openclaw | 4,255,537,377 | I_kwDOQb6kR879pljh | 65,978 | https://github.com/openclaw/openclaw/issues/65978 | https://api.github.com/repos/openclaw/openclaw/issues/65978 | Audio transcription skipped on direct delivery to idle agent (WhatsApp) | ## Description
When a voice note arrives on WhatsApp and the agent is **idle** (no active run), the message is delivered directly to the agent session **without** passing through `applyMediaUnderstanding`. The configured `tools.media.audio.models` (Deepgram nova-3, previously OpenAI gpt-4o-mini-transcribe) are never i... | closed | completed | true | 1 | [] | [] | 2026-04-13T14:17:29Z | 2026-04-28T04:56:23Z | 2026-04-25T17:31:00Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | alexandergirardet | 123,730,494 | U_kgDOB1_6Pg | User | false |
openclaw/openclaw | 4,255,591,214 | I_kwDOQb6kR879pysu | 65,981 | https://github.com/openclaw/openclaw/issues/65981 | https://api.github.com/repos/openclaw/openclaw/issues/65981 | [Bug]: isolated cron-agent runs cannot reliably use the OpenClaw cron tool, while main session can | ## Summary
OpenClaw `cron` tool works in the main session, but appears unavailable or mis-routed when a cron job itself runs as an **isolated cron agent** and tries to inspect cron state.
This breaks self-health-check jobs that are supposed to use the platform `cron` tool to inspect other cron jobs.
## Repro shape
... | closed | completed | true | 1 | [] | [] | 2026-04-13T14:24:39Z | 2026-04-16T04:24:15Z | 2026-04-13T22:50:45Z | NONE | null | 20260417T163222Z | 2026-04-17T16:32:22Z | aaajiao | 49,885 | MDQ6VXNlcjQ5ODg1 | User | false |
openclaw/openclaw | 4,255,610,311 | I_kwDOQb6kR879p3XH | 65,982 | https://github.com/openclaw/openclaw/issues/65982 | https://api.github.com/repos/openclaw/openclaw/issues/65982 | Multi-gateway: `--profile` flag routes restart/status to wrong LaunchAgent | ## Bug
When running multiple gateways via `--profile`, the `gateway restart` and `gateway status` commands resolve to the **wrong LaunchAgent plist**, causing cross-gateway restarts.
### Reproduction
```bash
# Setup: 3 gateways with profiles interactive, ops, batch
# Each has its own plist: ai.openclaw.interactive, ... | closed | completed | true | 1 | [] | [] | 2026-04-13T14:26:45Z | 2026-04-28T04:56:23Z | 2026-04-25T17:31:13Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | larrylhollan | 265,963,893 | U_kgDOD9pJdQ | User | false |
openclaw/openclaw | 4,255,615,976 | I_kwDOQb6kR879p4vo | 65,983 | https://github.com/openclaw/openclaw/issues/65983 | https://api.github.com/repos/openclaw/openclaw/issues/65983 | Background PTY exec runs can survive restart/session loss and become untracked orphan process trees | ## Summary
Background `exec` runs launched with `pty: true` appear to rely on in-memory-only ownership. If the gateway restarts or otherwise loses run state, the PTY-backed worker tree can remain alive while OpenClaw no longer knows about it. In practice this looks like orphaned Codex/OMX/MCP helper trees continuing t... | open | null | false | 1 | [] | [] | 2026-04-13T14:27:17Z | 2026-05-03T11:25:15Z | null | NONE | null | 20260503T120038Z | 2026-05-03T12:00:38Z | andydoucet | 65,342,358 | MDQ6VXNlcjY1MzQyMzU4 | User | false |
openclaw/openclaw | 4,255,654,750 | I_kwDOQb6kR879qCNe | 65,985 | https://github.com/openclaw/openclaw/issues/65985 | https://api.github.com/repos/openclaw/openclaw/issues/65985 | [Bug]: Telegram Active Memory slash command uses /active-memory, but BotFather-compatible command is /active_memory | ## Summary
There seems to be a Telegram slash-command naming mismatch for Active Memory.
The actual command exposed in chat is:
- `/active-memory`
But Telegram BotFather command registration expects a normal slash command token like:
- `/active_memory`
Because of that mismatch, the command users can actually add in... | closed | completed | true | 3 | [] | [] | 2026-04-13T14:30:48Z | 2026-04-27T04:49:30Z | 2026-04-24T22:29:45Z | NONE | null | 20260427T090020Z | 2026-04-27T09:00:20Z | aaajiao | 49,885 | MDQ6VXNlcjQ5ODg1 | User | false |
openclaw/openclaw | 4,255,724,294 | I_kwDOQb6kR879qTMG | 65,987 | https://github.com/openclaw/openclaw/issues/65987 | https://api.github.com/repos/openclaw/openclaw/issues/65987 | Task audit: inconsistent_timestamps (startedAt < createdAt) on cron-spawned tasks | ## Description
Cron-spawned tasks consistently have `startedAt` timestamps earlier than `createdAt` timestamps. This causes the tasks audit to report hundreds of `inconsistent_timestamps` warnings that are purely cosmetic — all tasks complete normally.
## Environment
- **OpenClaw version:** 2026.4.11 (769908e)
- **O... | closed | completed | true | 3 | [] | [] | 2026-04-13T14:37:58Z | 2026-04-30T04:40:54Z | 2026-04-26T05:02:33Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | AutonomIQ-Labs | 211,057,772 | U_kgDODJR8bA | User | false |
openclaw/openclaw | 4,255,767,956 | I_kwDOQb6kR879qd2U | 65,990 | https://github.com/openclaw/openclaw/issues/65990 | https://api.github.com/repos/openclaw/openclaw/issues/65990 | browser: managed Chrome fails with 'CDP websocket for profile openclaw is not reachable after start' while manual CDP works | ## Summary
OpenClaw managed browser fails to bring up a working CDP session for the default `openclaw` profile, even though the same Chrome binary and the same profile path work correctly when launched manually.
The managed browser path fails with:
`Error: Chrome CDP websocket for profile "openclaw" is not reachable... | closed | completed | true | 10 | [
"close:already-fixed",
"dedupe:child"
] | [
"vincentkoc"
] | 2026-04-13T14:44:46Z | 2026-04-28T05:21:01Z | 2026-04-25T11:41:36Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | clawdieclawdita | 259,446,051 | U_kgDOD3bVIw | User | false |
openclaw/openclaw | 4,255,784,342 | I_kwDOQb6kR879qh2W | 65,991 | https://github.com/openclaw/openclaw/issues/65991 | https://api.github.com/repos/openclaw/openclaw/issues/65991 | [Feature Request]: Add `silent` option to cron job `failureAlert` for Telegram delivery | ## Is your feature request related to a problem?
Cron job failure alerts currently always send with sound/notification on Telegram. For jobs that run overnight (healthchecks, monitoring), a failure alert at 03:00 will wake the user — even when the issue isn't urgent enough to warrant an immediate wake-up.
## Describe... | open | null | false | 1 | [] | [] | 2026-04-13T14:47:26Z | 2026-05-03T10:19:22Z | null | NONE | null | 20260503T120038Z | 2026-05-03T12:00:38Z | hugalafutro | 30,209,689 | MDQ6VXNlcjMwMjA5Njg5 | User | false |
openclaw/openclaw | 4,255,792,754 | I_kwDOQb6kR879qj5y | 65,993 | https://github.com/openclaw/openclaw/issues/65993 | https://api.github.com/repos/openclaw/openclaw/issues/65993 | Feishu streaming card produces duplicate/truncated messages with long replies | ## Bug Description
When using Feishu channel with `streaming: true` and `renderMode: "card"` (default configuration), long replies produce duplicate or truncated message blocks.
**Reproduction steps:**
1. Configure Feishu channel with streaming enabled
2. Send a request that generates a long reply (> ~1000 chars)
3. ... | closed | completed | true | 10 | [
"close:already-fixed",
"dedupe:child"
] | [
"vincentkoc"
] | 2026-04-13T14:48:52Z | 2026-04-30T04:47:39Z | 2026-04-25T05:52:24Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | Dany-Deng-cn | 273,966,579 | U_kgDOEFRl8w | User | false |
openclaw/openclaw | 4,255,802,462 | I_kwDOQb6kR879qmRe | 65,994 | https://github.com/openclaw/openclaw/issues/65994 | https://api.github.com/repos/openclaw/openclaw/issues/65994 | [Bug]: exec completion events leak into webchat conversation as System messages | ## Summary
Async exec completion events are currently routed through the heartbeat/alert path and end up being rendered into the webchat/control-ui conversation as visible System messages.
This pollutes the user-facing transcript with internal runtime notifications that should not appear as normal chat messages.
## ... | closed | not_planned | true | 7 | [
"bug",
"bug:behavior"
] | [] | 2026-04-13T14:50:28Z | 2026-04-30T04:40:53Z | 2026-04-26T05:02:37Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | EiYaFi | 105,110,492 | U_kgDOBkPb3A | User | false |
openclaw/openclaw | 4,255,892,848 | I_kwDOQb6kR879q8Vw | 65,997 | https://github.com/openclaw/openclaw/issues/65997 | https://api.github.com/repos/openclaw/openclaw/issues/65997 | plugin allowlist validation conflicts with CLI gating for help / plugin in OpenClaw 2026.4.12 | I ran into an inconsistency in OpenClaw `2026.4.12` around `plugins.allow`.
## Summary
When `plugins.allow` is set to a narrow allowlist, some CLI surfaces become unavailable unless `"help"` and `"plugin"` are added to the allowlist.
However, when adding those entries, config validation warns that they are not real ... | closed | duplicate | true | 2 | [
"duplicate",
"dedupe:child"
] | [] | 2026-04-13T15:04:33Z | 2026-04-28T04:36:35Z | 2026-04-26T03:54:54Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | jinpeng | 347,816 | MDQ6VXNlcjM0NzgxNg== | User | false |
openclaw/openclaw | 4,255,904,834 | I_kwDOQb6kR879q_RC | 65,998 | https://github.com/openclaw/openclaw/issues/65998 | https://api.github.com/repos/openclaw/openclaw/issues/65998 | active-memory: toolsAllow filter does not exclude MCP tool schemas from sub-agent API call, causing 400 errors on large MCP deployments | ## Summary
The `active-memory` plugin's `toolsAllow: ["memory_search", "memory_get"]` parameter correctly limits which tools the sub-agent **can call**, but it does **not** prevent other MCP servers' tool schemas from being sent to the model API. On deployments with large MCP servers (e.g., ms365 with 100+ tools), thi... | closed | completed | true | 2 | [] | [] | 2026-04-13T15:06:17Z | 2026-04-30T04:40:52Z | 2026-04-26T05:02:41Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | darrinwillard | 267,111,867 | U_kgDOD-vNuw | User | false |
openclaw/openclaw | 4,255,951,349 | I_kwDOQb6kR879rKn1 | 66,003 | https://github.com/openclaw/openclaw/issues/66003 | https://api.github.com/repos/openclaw/openclaw/issues/66003 | memory-wiki: per-agent vault path isolation for multi-tenant deployments | ## Context
Multi-tenant SaaS deployments running OpenClaw need per-tenant (per-org) isolation of the memory-wiki vault. The current `plugins.entries.memory-wiki.config.vault.path` is a static string resolved once at plugin registration, shared across every agent on the gateway process.
## The gap
Per source inspecti... | closed | not_planned | true | 3 | [] | [] | 2026-04-13T15:13:19Z | 2026-04-30T04:40:52Z | 2026-04-26T05:02:45Z | NONE | null | 20260430T060018Z | 2026-04-30T06:00:18Z | yoanndefay | 65,121,829 | MDQ6VXNlcjY1MTIxODI5 | User | false |
openclaw/openclaw | 4,255,952,005 | I_kwDOQb6kR879rKyF | 66,004 | https://github.com/openclaw/openclaw/issues/66004 | https://api.github.com/repos/openclaw/openclaw/issues/66004 | [Bug]: subagent with timeoutSeconds>0 returns ok but frozenResultText null, causes Cannot read properties of undefined | ### Bug type
Crash (process/app exits or hangs)
### Beta release blocker
No
### Summary
subagent with timeoutSeconds>0 returns status:ok but frozenResultText null, causes Cannot read properties of undefined
### Steps to reproduce
1. Set agents.defaults.timeoutSeconds: 600 in openclaw.json
2. Restart gateway
3. S... | closed | completed | true | 1 | [
"bug"
] | [] | 2026-04-13T15:13:26Z | 2026-04-28T04:56:22Z | 2026-04-25T17:31:20Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | Salamander-c | 270,838,582 | U_kgDOECSrNg | User | false |
openclaw/openclaw | 4,255,955,280 | I_kwDOQb6kR879rLlQ | 66,006 | https://github.com/openclaw/openclaw/issues/66006 | https://api.github.com/repos/openclaw/openclaw/issues/66006 | Webchat: Auto-scroll to latest message not working | ## Bug Description
In the Webchat interface, the chat does not automatically scroll to the latest message after a reply is generated. The user must manually scroll down using the mouse wheel or keyboard to see the new content.
## Steps to Reproduce
1. Open Webchat interface
2. Send a message to the agent
3. Wait for... | closed | completed | true | 1 | [] | [] | 2026-04-13T15:13:54Z | 2026-04-28T04:56:21Z | 2026-04-25T17:31:26Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | chenhuihui8429 | 273,324,543 | U_kgDOEEqZ_w | User | false |
openclaw/openclaw | 4,255,988,163 | I_kwDOQb6kR879rTnD | 66,008 | https://github.com/openclaw/openclaw/issues/66008 | https://api.github.com/repos/openclaw/openclaw/issues/66008 | [Bug]: Thinking lines are missing whitespace | ### Bug type
Behavior bug (incorrect output/state without crash)
### Beta release blocker
No
### Summary
Streamed thinking output is missing whitespace and sentences just get smushed together.
### Steps to reproduce
1. Use TUI and ask a question
2. Observe streamed thinking output.
### Expected behavior
Stream... | closed | completed | true | 2 | [
"bug",
"bug:behavior"
] | [] | 2026-04-13T15:19:02Z | 2026-04-28T04:56:20Z | 2026-04-25T17:31:40Z | NONE | null | 20260428T060020Z | 2026-04-28T06:00:20Z | AugustSaintFreytag | 7,656,669 | MDQ6VXNlcjc2NTY2Njk= | User | false |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.