#I'm trying to use openclaw to talk to my openclaw agent and this message keeps coming up:

1 messages · Page 1 of 1 (latest)

stray dew
#

OpenClaw
Gateway Dashboard
WebSocket URL
ws://127.0.0.1:18789
Gateway Token
••••••••••••••••••••••••••••••••••••••••••••••••

Password (not stored)
optional

Connect
disconnected (1006): no reason

crisp tendonBOT
#

To help others find answers, you can mark your question as solved via Right click solution message -> Apps -> ✅ Mark Solution

wraith doveBOT
#

Bootstrap complete — no recent resume file/highlights found, so I’m treating this as a fresh support thread.

For this dashboard disconnected (1006): no reason, first please paste redacted output from:

openclaw --version
openclaw gateway status
openclaw status
openclaw doctor

Most likely checks:

  1. Run:
# 
openclaw dashboard

and use the URL it prints instead of manually entering ws://127.0.0.1:18789; the dashboard command opens the Control UI with the current auth context. (docs.openclaw.ai)

  1. If the gateway is on a different machine/VPS, 127.0.0.1 points at your browser’s machine, not the gateway. Use the tunnel/Tailscale/remote URL instead.

  2. If it still disconnects, run:

openclaw logs --follow

and look for auth/origin/gateway errors. The docs’ dashboard-connect triage is: check gateway running, health, correct token, remote tunnel/Tailscale if applicable, then tail logs. (docs.openclaw.ai)

Don’t paste your token publicly — redacted logs are perfect.

stray dew
#

bash-3.2$ openclaw logs --follow

🦞 OpenClaw 2026.4.24 (cbcfdf6) — Ah, the fruit tree company! 🍎

Gateway not reachable. Is it running and accessible?
Gateway target: ws://127.0.0.1:18789
Source: local loopback
Config: /Users/kevinward/.openclaw/openclaw.json
Bind: lan

wraith doveBOT
#

That output means the dashboard isn’t the issue yet — the Gateway is not reachable/running on 127.0.0.1:18789.

Try this:

openclaw gateway status
openclaw gateway start
openclaw gateway status
openclaw dashboard

If gateway start says no service is installed, do:

openclaw gateway install --force
openclaw gateway start
#

If it starts and immediately dies, run it in the foreground so we can see the real error:

openclaw gateway run --verbose

Also, since openclaw logs --follow can’t reach the gateway, use the fallback log tail:

tail -f "$(ls -t /tmp/openclaw/openclaw-*.log | head -1)"

Paste the gateway status output or the first foreground error here, with tokens redacted.

Docs: https://docs.openclaw.ai/gateway/troubleshooting

stray dew
#

Runtime: running (pid 84707, state active)
Warm-up: launch agents can take a few seconds. Try again shortly.
Connectivity probe: failed
Probe target: ws://127.0.0.1:18789
connect ECONNREFUSED 127.0.0.1:18789
Capability: unknown

Gateway port 18789 is not listening (service appears running).
Last gateway error: 2026-04-17T18:07:37.074-06:00 Gateway start blocked: existing config is missing gateway.mode. Treat this as suspicious or clobbered config. Re-run openclaw onboard --mode local or openclaw setup, set gateway.mode=local manually, or pass --allow-unconfigured.
Logs: ~/.openclaw/logs/gateway.log
Errors: ~/.openclaw/logs/gateway.err.log
Restart log: ~/.openclaw/logs/gateway-restart.log

Troubles: run openclaw status
Troubleshooting: https://docs.openclaw.ai/troubleshooting
bash-3.2$

wraith doveBOT
#

Nice, we found the actual blocker: your config is missing gateway.mode, so the service exists but the gateway refuses to start — that’s why the dashboard gets 1006.

Safest fix:

cp ~/.openclaw/openclaw.json ~/.openclaw/openclaw.json.bak-$(date +%Y%m%d-%H%M%S)

openclaw onboard --mode local
openclaw config validate
openclaw gateway restart
openclaw gateway status
openclaw dashboard

If you want the smallest manual repair instead: