#chain-error
Sorted by views, then solution_desc, solution, and root_cause length (desc).
2584 issues
OpenClaw 2026.3.28 fails on every chat turn for a custom OpenAI-compatible provider with “LLM request failed: network connection error”, even though direct /v1/models and /v1/chat/completions calls to the same endpoint succeed from the same Windows machine.
The OpenClaw Gateway service fails to install with a systemd error indicating the service unit file does not exist. This affects users attempting to run the gateway as a systemd user service.
When attempting to send outbound WhatsApp messages via `openclaw message send` or `sessions_send` RPC tool, the gateway returns `UNAVAILABLE: No active WhatsApp Web listener (account: default)` — even though `openclaw channels status` shows `linked, running, connected` and inbound messages + auto-replies work perfectly.
When multiple model providers (e.g., Azure OpenAI and Zhipu) are configured in OpenClaw, switching to a model from a non-default provider in the UI does not correctly update the routing. The request is still sent to the default provider, resulting in a "model not allowed" error.
Sending any image via Telegram results in a generic error response: **"Something went wrong while processing your request. Please try again."** This began after updating from **2026.3.31 → 2026.4.2**. Images were working prior to the update.
When trying to install a community plugin (`openclaw-codex-app-server`) that legitimately uses `child_process` (to spawn the Codex CLI), installation is blocked with no way to proceed: \`\`\` WARNING: Plugin "openclaw-codex-app-server" contains dangerous code patterns: Shell command execution detected (child_process) (src/client.ts:660) Plugin "openclaw-codex-app-server" installation blocked: dangerous code patterns detected \`\`\`
On this OpenClaw 2026.4.2 Linux host, the official managed browser lane for profile `openclaw` is degraded even though Chromium/CDP itself is healthy.
When a session context becomes large enough to trigger compaction, and that compaction process times out, the agent enters a silent failure loop. Each timeout (~10 min) triggers a delivery retry that resends the same message to the user repeatedly — with no recovery, no fallback, and no way for the session to self-resolve.
$ pnpm install Scope: all 71 workspace projects Lockfile is up to date, resolution step is skipped Packages: +1285 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ WARN GET https://codeload.github.com/whiskeysockets/libsignal-node/tar.gz/1c30d7d7e76a3b0aa120b04dc6a26f5a12dccf67 error (ECONNRESET). Will retry in 10 seconds. 2 retries left. WARN GET https://codeload.github.com/tloncorp/api-beta/tar.gz/7eede1c1a756977b09f96aa14a92e2b06318ae87 error (ECONNRESET). Will retry in 10 seconds. 2 retries left. WARN GET https://codeload.github.com/whiskeysockets/libsignal-node/tar.gz/1c30d7d7e76a3b0aa120b04dc6a26f5a12dccf67 error (ECONNRESET). Will retry in 1 minute. 1 retries left. WARN GET https://codeload.github.com/tloncorp/api-beta/tar.gz/7eede1c1a756977b09f96aa14a92e2b06318ae87 error (ECONNRESET). Will retry in 1 minute. 1 retries left. Progress: resolved 1285, reused 1283, downloaded 0, added 1283
**Version:** 2026.3.13 (updated 2026-03-16) **OS:** Linux 6.8.0-101-generic (x64)
When running `openclaw gateway restart`, the Gateway service stops but fails to restart. Only a manual system restart (or computer reboot) restores the service.
When the agent sends a reply via iMessage, the sent message is picked up by `imsg watch` and fed back to the agent as a new inbound message from the remote party. This creates echo loops where the agent sees its own replies as user messages.