Health check failed: gateway closed (1006 abnormal closure (no close frame)): no close reason

[coral rune]

@hot star

[hot star]

⚠️ Agent failed before reply: No API key found for provider "anthropic". Auth store: /srv/clawdbot/state/agents/main/agent/auth-profiles.json (agentDir: /srv/clawdbot/state/agents/main/agent). Configure auth for this agent (clawdbot agents add <id>) or copy auth-profiles.json from the main agentDir.
Logs: clawdbot logs --follow

[final topaz]

openclaw models status

[agile wharf]

hey mate were you able to figure this out?

[hot star]

Not 100% yet — but that specific 1006 “abnormal closure (no close frame)” is usually not an auth problem; it’s the WebSocket getting dropped (gateway restarting/crashing, or a proxy/tunnel/firewall in between killing WS).

To narrow it down fast, can you paste the output of these from the gateway host:

openclaw --version
openclaw gateway status
openclaw status
openclaw doctor
openclaw logs --follow

Also: are you connecting to the Control UI locally (same machine) or remotely (VPS / Cloudflare / nginx / Tailscale Serve)?

Quick split that often finds the culprit:

• If openclaw gateway status shows RPC probe: ok but the Control UI still drops with 1006 → it’s usually browser/proxy/tunnel WS config.
• If logs show the gateway process exiting/restarting around the same time → it’s a gateway crash / supervisor restart (the openclaw logs --follow snippet will show it).

If you tell me “local vs remote” + paste those outputs (even just the relevant log chunk around the disconnect), I can point to the exact fix.