Onboarding failure

[glad geode]

I onboarded my openclaw entered my oauth for codex and when i finish the process it says Onboarding complete. Dashboard Opened but the dashboard says this site cant be reached

[native owl]

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

[main tiger]

That “site can’t be reached” usually means the Gateway (which serves the dashboard) isn’t reachable on the URL the onboarding opened.

A few quick checks to pinpoint it:

1. Paste the exact dashboard URL your browser tried to open (just the host/port part is fine).
2. Are you running OpenClaw on the same machine as the browser, or is it on a remote box / Docker / SSH / VPS?

Then please run these and paste the output:

openclaw --version
openclaw gateway status
openclaw status --all
openclaw dashboard --no-open

Notes / common fixes:

• The dashboard is normally at http://127.0.0.1:18789/ (served by the running Gateway). If nothing is listening there, the browser will show “can’t be reached”.
• If you onboarded on a remote machine, 127.0.0.1:18789 is remote localhost, so your local browser can’t reach it. Use an SSH tunnel:

  ssh -N -L 18789:127.0.0.1:18789 user@YOUR_HOST
  

  then open the tokenized URL from openclaw dashboard --no-open.

Docs reference (matches the above): https://docs.openclaw.ai/web/dashboard and https://docs.openclaw.ai/gateway/troubleshooting

Two things:

1) You pasted your Gateway dashboard token publicly

That token is basically “admin access” to your Control UI. Please delete/edit that message if you can, and rotate the token.

On your machine, run:

openclaw config set gateway.auth.token "<new-random-token>"
openclaw gateway restart
openclaw dashboard --no-open

(Then use the new link it prints. Don’t post the token again.)

Docs: https://docs.openclaw.ai/web/dashboard

2) “This site can’t be reached” = nothing is listening on 127.0.0.1:18789

So the Gateway likely isn’t running (or crashed / didn’t bind).

Please paste output of:

openclaw gateway status
openclaw status --all

And also check whether the port is listening:

netstat -ano | findstr :18789

What I’m looking for:

• If openclaw gateway status shows not running → run openclaw gateway start (or openclaw gateway restart) and try http://127.0.0.1:18789/ again.
• If it says running but netstat shows nothing listening on :18789 → the logs/status output will usually say why (bind/auth/config error, port in use, etc.).

Troubleshooting reference: https://docs.openclaw.ai/gateway/troubleshooting

[glad geode]

🦞 OpenClaw 2026.2.6-3 (85ed6c7) — Siri's competent cousin.

|
o
Service: Scheduled Task (registered)
File logs: \tmp\openclaw\openclaw-2026-02-07.log

Config (cli): C:\Users\User.openclaw\openclaw.json
Config (service): C:\Users\User.openclaw\openclaw.json

Gateway: bind=loopback (127.0.0.1), port=18789 (env/config)
Probe target: ws://127.0.0.1:18789
Dashboard: http://127.0.0.1:18789/
Probe note: Loopback-only gateway; only local clients can connect.

Runtime: stopped (state Queued, last run time 2/7/2026 2:04:37 PM)
RPC probe: failed
RPC target: ws://127.0.0.1:18789
gateway closed (1006 abnormal closure (no close frame)): no close reason
Gateway target: ws://127.0.0.1:18789
Source: cli --url
Config: C:\Users\User.openclaw\openclaw.json

Service is loaded but not running (likely exited immediately).
File logs: \tmp\openclaw\openclaw-2026-02-07.log
Logs: schtasks /Query /TN "OpenClaw Gateway" /V /FO LIST

[main tiger]

Your openclaw gateway status is the smoking gun: the Gateway isn’t running (Runtime: stopped). That’s why http://127.0.0.1:18789/ says “site can’t be reached” — nothing is listening.

1) Try starting it

Run:

openclaw gateway start
openclaw gateway status

If it flips to Runtime: running, refresh http://127.0.0.1:18789/.

2) If it immediately stops again, run it in the foreground to see the real error

This usually shows the exact crash reason (missing runtime, port in use, config issue, etc.):

openclaw gateway run --verbose

Copy/paste whatever error it prints when it exits.

3) Also paste the last ~100 lines of the file log

Your status says logs are here:

• \tmp\openclaw\openclaw-2026-02-07.log (typically C:\tmp\openclaw\...)

PowerShell:

Get-Content \tmp\openclaw\openclaw-2026-02-07.log -Tail 120

4) Quick port-collision check

netstat -ano | findstr :18789

One more important thing: you posted the dashboard token earlier — please rotate gateway.auth.token after we get the Gateway running (and don’t paste the new one publicly).

[glad geode]

13:14:27 [ws] unauthorized conn=2a59da1b-624d-4872-ac6b-d16ffef2aedf remote=127.0.0.1 client=openclaw-control-ui webchat vdev reason=token_missing
13:14:27 [ws] closed before connect conn=2a59da1b-624d-4872-ac6b-d16ffef2aedf remote=127.0.0.1 fwd=n/a origin=http://127.0.0.1:18789 host=127.0.0.1:18789 ua=Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/144.0.0.0 Safari/537.36 code=4008 reason=connect failed
13:14:43 [ws] unauthorized conn=c2a4bd3d-9d8d-4f3d-8c68-b83535645775 remote=127.0.0.1 client=openclaw-control-ui webchat vdev reason=token_missing
13:14:43 [ws] closed before connect conn=c2a4bd3d-9d8d-4f3d-8c68-b83535645775 remote=127.0.0.1 fwd=n/a origin=http://127.0.0.1:18789 host=127.0.0.1:18789 ua=Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/144.0.0.0 Safari/537.36 code=1008 reason=unauthorized: gateway token missing (open the dashboard URL and paste the token in Control UI settings)

[main tiger]

Good news: the Gateway is up now — the log is just saying the Control UI is connecting without a token (reason=token_missing), so the WebSocket handshake gets rejected.

Do this on the same machine:

1. Get a fresh tokenized dashboard link:

openclaw dashboard --no-open

2. Open the URL it prints (it should be http://127.0.0.1:18789/?token=...).

• Don’t open plain http://127.0.0.1:18789/ without the token.
• In your earlier message you had #token=... in the URL — if that’s what you’re using, grab a fresh link from openclaw dashboard (current docs expect ?token=...).

3. If you already have the dashboard open: open Settings in the Control UI and paste the token there (it stores it in localStorage), then reload.

One important thing: you pasted the token publicly earlier — rotate it when you can: