API middleware

Browser cannot attach/use healthy local Brave CDP on hardened Debian 13 VPS

On a hardened Debian 13 VPS, OpenClaw browser automation fails even though Brave is running headless, exposes a valid CDP endpoint on localhost, and returns a valid `webSocketDebuggerUrl`. OpenClaw reports the browser profile as **not running**, returns **no tabs**, and `browser open` fails with **PortInUseError** instead of attaching to the existing CDP browser. This appears to be an OpenClaw Linux/VPS browser attach/runtime issue, not a missing-browser or broken-CDP issue.

why the ollama get response very slow? (QWEN3.5 35B A3B)

[Bug]: Local Ollama agent runs intermittently fail with 'fetch failed' / timeout on 2026.3.8-beta.1

On OpenClaw `2026.3.8-beta.1`, local Ollama models are callable directly via `ollama run`, but OpenClaw agent runs intermittently fail with: - `LLM request timed out` - `rawErrorPreview":"fetch failed"` This happens for both: - `qwen3.5:122b-a10b` - `gpt-oss:120b`

[BUG] claude.ai Gmail MCP auth flow broken in Claude Code CLI on Linux — Enter key does nothing after browser auth

bug(agents): OpenAI WS replay drops message-level phase when block has explicit textSignature — commentary leaks on history replay

[Bug]: The Gateway daemon of OpenClaw crashes on the laptop when the device switches from being powered by the AC adapter to battery power.

CPU Intel(R) Core(TM) i7-4900MQ CPU @ 2.80GHz 2.80 GHz RAM 32.0 GB (31.7 GB 可用) 238 GB SSD SAMSUNG MZ7LN256HCHP-000L7, 1.82 TB HDD ST2000LM015-2E8174 NVIDIA Quadro K3100M (4 GB), Intel(R) HD Graphics 4600 (113 MB) Windows 10 专业版 VERSION:22H2 OS Internal version 19045.6466 I normally use the laptop with the AC power adapter plugged in, and OpenClaw runs completely fine. But as soon as I unplug the adapter and switch to battery power, the OpenClaw Gateway daemon crashes every single time — the failure rate is 100%.

[Bug/UX] Control UI loses gateway token after container rebuild

langchain-ollama: Error with using local model `gwen3:8b`

Encountered this error although `gwen3:8b` has been pulled by ollama (and indicated in the error log too)

MiniMax M2.7 Cloud model request

[Bug]: Channels (Telegram/WhatsApp) not loading in v2026.4.2 - empty table in status

Channels Empty

[Bug]: UI/Interface

Web UI model selector sends bare model IDs with wrong provider prefix when multiple providers are configured, causing "model not allowed" errors for non-default providers.

CUDA error on RTX 3060 Laptop 6GB with qwen3.5:4b — fixed by updating from 0.17.1 to 0.17.7

Cron: gateway restart loop should pause/defer scheduled jobs

[Bug] Gateway restart fails - stale process misdetection

Gateway restart fails due to stale process misdetection and race condition

[Bug]: Gemma4 on vLLM + PI coding agent: Validation failed for tool "edit": - path: must have required property 'path'

[Feature]: Add Rotorquant support

[Bug]: Windows： Gateway 关闭 PowerShell 窗口后断连

在 Windows 上使用 `openclaw gateway install` 安装的计划任务，在关闭 PowerShell 窗口后 Gateway 会断连，必须重新运行 `openclaw gateway start` 才能恢复。

v2026.4.5: Missing runtime dependencies cause CLI crash on fresh global install

A fresh `npm install -g openclaw` of v2026.4.5 crashes immediately on any command (e.g. `openclaw doctor`) because 8 packages are imported in the bundled dist files but not declared in `package.json` `dependencies`. This affects anyone doing a clean global install or update. It is **not** environment-specific — npm only installs declared dependencies, so the missing packages are never present.

Slack Socket Mode: event loop starvation causes pong timeouts and silent message loss

Investigate Gemini rate limiting in Matrix/agent workflows

Serena triage memory already tracked Gemini rate limiting as an operational issue before the current incident.