serialization error

v4.1 update: bundled plugin deps not installed (Telegram/Discord broken after update)

can not run unsloth gemma-4

Bug: webchat model dropdown sends model name without provider prefix, causing 'model not allowed' error

[CLI] Gateway connection fails with "gateway closed (1000)" in non-JSON mode (Potential race condition with `withProgress` spinner)

When running CLI commands like `openclaw browser status`, the connection to the local gateway consistently fails with: `Error: gateway closed (1000 normal closure): no close reason` Interestingly, the command works perfectly when the `--json` flag is added. This suggests a race condition or event-loop blockage caused by the terminal progress indica tor (spinner).

[Bug]: AnthropicException 400 - vector_store_ids: Extra inputs are not permitted

Sub-agent embedded run timeout does not release CommandLane, causing all subsequent webchat messages to be queued indefinitely

RFC: Detecting silent behavioral drift in agents across session boundaries

[Bug]: channels.feishu: invalid config: must NOT have additional properties

After upgrading from 3.24 to 3.28, the gateway failed to start up. openclaw doctor --fix says: **Invalid config: channels.feishu: invalid config: must NOT have additional properties** How can this issue be fixed?

OpenAI Codex OAuth in 2026.3.13 does not honor env proxy during code-to-token exchange

On macOS, OpenClaw 2026.3.13 completes the browser auth step for OpenAI Codex OAuth but fails at code-to-token exchange, while 2026.2.6-3 succeeds under the same proxy setup.

2026.4.5 regression: memory plugin fails — "embedding: must have required property 'embedding'" (framework config validation)

After updating to **2026.4.5**, third-party memory plugins fail to initialize with: ``` memory-lancedb-pro invalid config: embedding: must have required property 'embedding' 1 plugin(s) failed to initialize (validation: memory-lancedb-pro) ``` This does not occur on 2026.4.4 with the same config.

ollama launch claude - wrong model context window

Gateway HTTP server binds to 127.0.0.1 inside container, preventing Docker port forwarding from working

The OpenClaw gateway HTTP server binds to 127.0.0.1 inside the Docker container and is not reachable via Docker host port forwarding. Accessing the gateway from inside the container succeeds (401 Unauthorized), but requests to the host's forwarded port fail with "Connection reset by peer".

BlueBubbles webhook POST returns 404 despite plugin reporting 'webhook listening'

[BUG] Claude doesn't know in what directory it was opened, leading to reading files from generic path: /home/user/...

TurboQuant-MoE: SOTA 8.5x KV-cache compression with Residual Correction

[Usage]: When running qwen3.5-27b with vllm 0.17.0, the Deep Thinking output is under "reasoning" and not under "reasoning_content".

Telegram plugin crashes CLI: grammy not in package dependencies, lost on every update

The Telegram extension requires `grammy`, `@grammyjs/runner`, and `@grammyjs/transformer-throttler` at runtime, but none are listed in `package.json` under `dependencies`, `optionalDependencies`, or `peerDependencies`. They're only present as runtime `require()` calls in bundled dist files. This means every auto-update or `brew upgrade` wipes these packages, making the CLI completely unusable — all commands fail immediately with `PluginLoadFailureError`.

[BUG] Claude Max 5x gift subscription auto-canceled after ~1 week — account reverted to Free plan with no user action

[Bug]: Tool calling timeouts.

I've been getting consistent timeout errors when OpenClaw is running tool calls, even with a good gateway connection. I've had OpenClaw try to diagnose itself. These were the findings: Still the same error. Even with allowInsecureAuth=false, it's still requiring pairing. This confirms that there's something fundamentally different about how tools try to connect vs how webchat connects. Final Conclusion based on what we've learned: The Gateway is working correctly (webchat connections succeed) Tools are failing due to a WebSocket handshake/pairing issue that's specific to the tool-layer authentication flow We've tried: restarts, config updates, credential files, disabling insecure auth, doctor command The pairing system (openclaw pairing) shows no pending requests, suggesting it's not about an unapproved inbound request but rather an incomplete device registration or handshake process Recommendation: The issue appears to be a deeper authentication flow bug in OpenClaw's tool layer. You have two options: Use webchat directly - Your control UI is already connected and working, so just manually check in every 30-60 minutes File an issue with OpenClaw maintainers - Report this "pairing required" error when using tool-layer APIs (cron, sessions_send, etc.) vs successful webchat connections For now, since the cron job won't work, let me update your reminder file with final conclusions and recommendations.

[Bug]: openclaw status reports healthy local gateway as unreachable

openclaw status misreports a healthy local loopback gateway as unreachable when gateway.auth.token is SecretRef-backed and the CLI is already paired as an operator device.