#onboarding issues

1 messages · Page 1 of 1 (latest)

bold kelpBOT
#

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

tired basinBOT
#

Perfect — please paste these first (in this order):

  1. openclaw --version
  2. openclaw status --all
  3. openclaw gateway status
  4. Reproduce each error once, then paste ~40 lines around it from:
    openclaw logs --follow

Also include:

  • your OS (macOS/Linux/WSL/Windows)
  • how you installed (installer script / npm / git clone)
  • whether “restricted access” is in Dashboard UI, CLI, or channel replies

If you can, split logs into two sections:

  • A: “Cannot find module”
  • B: “Gateway bound but restricted access”
rancid mist
tired basinBOT
#

Nice, this helps a lot.

From your paste: the main issue right now is gateway service is stopped/exiting immediately (Task Last Run Result=1), so logs --follow can’t attach yet.

Try this exact recovery sequence:

openclaw doctor
openclaw gateway install --force
openclaw gateway start
openclaw gateway status

Then paste the new openclaw gateway status output.

If it still exits, paste these two outputs next (this is where the “Cannot find module” line usually appears):

# 
type "%LOCALAPPDATA%\Temp\openclaw\openclaw-2026-04-07.log"
schtasks /Query /TN "OpenClaw Gateway" /V /FO LIST

Quick note on your “restricted access” symptom:

  • your gateway is currently bind=loopback, so only local clients can reach it.
  • and auth may still require token even on localhost.

After service is up, if Dashboard says unauthorized:

openclaw config get gateway.auth.token

then paste that token into Control UI settings.

Refs: https://docs.openclaw.ai/help/faq and https://docs.openclaw.ai/gateway/troubleshooting

rancid mist
#

C:\Windows\System32>openclaw gateway status

🦞 OpenClaw 2026.4.7 (5050017) — Half butler, half debugger, full crustacean.

|
o
Service: Scheduled Task (registered)
File logs: ~\AppData\Local\Temp\openclaw\openclaw-2026-04-07.log
Command: C:\Program Files\nodejs\node.exe C:\Users\leela\AppData\Roaming\npm\node_modules\openclaw\dist\index.js gateway --port 18789
Service file: ~.openclaw\gateway.cmd
Service env: OPENCLAW_GATEWAY_PORT=18789

Config (cli): ~.openclaw\openclaw.json
Config (service): ~.openclaw\openclaw.json

Gateway: bind=loopback (127.0.0.1), port=18789 (service args)
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: running (last run 267009, last run time 4/7/2026 11:36:58 PM)
RPC probe: ok

Listening: 127.0.0.1:18789
Troubles: run openclaw status
Troubleshooting: https://docs.openclaw.ai/troubleshooting

C:\Windows\System32>run openclaw status
'run' is not recognized as an internal or external command,
operable program or batch file.

C:\Windows\System32>openclaw gateway start

🦞 OpenClaw 2026.4.7 (5050017) — I read logs so you can keep pretending you don't have to.

Restarted Scheduled Task: OpenClaw Gateway

C:\Windows\System32>openclaw config get gateway.auth.token

🦞 OpenClaw 2026.4.7 (5050017) — I've read more man pages than any human should—so you don't have to.

OPENCLAW_REDACTED

tired basinBOT
#

Awesome — this is healthy now ✅

  • Runtime: running
  • RPC probe: ok
  • Listening: 127.0.0.1:18789

So gateway is up.

Also, tiny gotcha:
Troubles: run openclaw status means “run this command”, not a literal run command.
Use just:

openclaw status

For your restricted access symptom, this is likely expected with current config:

#
  • bind=loopback = only this same machine can connect
  • token auth is enabled, so dashboard needs the token you fetched

If you’re connecting locally, open http://127.0.0.1:18789/ and paste the token in Gateway Access.

If you want access from another device, switch to remote-safe setup (Tailscale/LAN + auth) — I can walk you through that next.

If “Cannot find module” appears again, paste:

  1. exact error line
  2. where openclaw
  3. type "%LOCALAPPDATA%\\Temp\\openclaw\\openclaw-YYYY-MM-DD.log" (around that timestamp)
rancid mist
#

I get internal server error

tired basinBOT
#

Got it — “internal server error” is usually fixable fast.

First, tell me where you see it (Dashboard page, Control UI connect, or CLI command).

Then run this:

openclaw gateway restart
openclaw status --all

And while reproducing the error, keep this running in another terminal:

openclaw logs --follow
#

Paste ~40 lines around the error.

If this is in Dashboard/Control UI specifically:

  1. use http://127.0.0.1:18789/
  2. in Gateway Access, use token (leave password empty unless you explicitly configured password mode)
  3. paste token again from openclaw config get gateway.auth.token

Don’t share the actual token value in chat.

rancid mist
#

23:47:45-05:00 [ws] closed before connect conn=e005f642-d6bd-47bc-bc5d-e92326ff2b01 remote=127.0.0.1 fwd=n/a origin=n/a host=127.0.0.1:18789 ua=n/a code=1008 reason=connect failed

tired basinBOT
#

Nice catch — that log is a WebSocket handshake/auth issue, not a module crash.

code=1008 reason=connect failed usually means a client connected to the gateway port but didn’t complete the required connect auth handshake (wrong client, stale UI state, or missing token).

Try this exact flow:

openclaw gateway restart
openclaw dashboard

Then:

  • open the dashboard URL it prints (http://127.0.0.1:18789/)
  • paste token from openclaw config get gateway.auth.token in Gateway Access
  • do a hard refresh (or try Incognito) if it still says internal server error

Also make sure you’re not opening ws://127.0.0.1:18789 directly in a browser tab.

#

If it still fails, paste 20–30 log lines before and after that 1008 line (there’s usually an auth detail code nearby).

rancid mist
#

but my token is getting redacted on running command.

tired basinBOT
#

Yep — that’s expected. openclaw config get gateway.auth.token is intentionally redacted for safety.

Use one of these instead:

  1. Easiest: just run
openclaw dashboard

It should bootstrap auth automatically.

  1. If UI still asks for token, open config file directly (local only, don’t share it):
notepad %USERPROFILE%\.openclaw\openclaw.json

Then copy gateway.auth.token from there into Gateway Access.

Also, don’t paste that token in Discord.

rancid mist
tired basinBOT
#

Got it — if openclaw dashboard gives a #token=... URL and you still get Internal Server Error, this is now a server-side route error (not a token visibility problem).

Do this:

openclaw doctor
openclaw gateway restart

Then immediately collect the failing stack:

  1. Run:
openclaw logs --follow
  1. Refresh http://127.0.0.1:18789/#token=... once
  2. Paste the error lines that appear right after refresh
#

That log snippet will show the exact root cause (usually a missing module/control-ui asset or a thrown route error), and I can give you the exact fix next.
(keep token hidden/redacted)

rancid mist
#

I have setup anthropic api key and setup claude 4.6 sonnet

⚠️ Agent failed before reply: All models failed (2): opencode/claude-sonnet-4-6: No API key found for provider "opencode". Auth store: C:\Users\leela.openclaw\agents\main\agent\auth-profiles.json (agentDir: C:\Users\leela.openclaw\agents\main\agent). Configure auth for this agent (openclaw agents add <id>) or copy auth-profiles.json from the main agentDir. (auth) | openai/gpt-5.4: No API key found for provider "openai". You are authenticated with OpenAI Codex OAuth. Use openai-codex/gpt-5.4 (OAuth) or set OPENAI_API_KEY to use openai/gpt-5.4. (auth).
Logs: openclaw logs --follow

yet this

tired basinBOT
#

Yep — this error explains it clearly: your model config is pointing to the wrong providers.

You currently have:

  • opencode/claude-sonnet-4-6 → provider opencode (no key configured)
  • fallback openai/gpt-5.4 → requires OPENAI_API_KEY (Codex OAuth does not satisfy this)

Run this:

openclaw models set anthropic/claude-sonnet-4-6
openclaw models fallbacks remove openai/gpt-5.4
openclaw models fallbacks add openai-codex/gpt-5.4
openclaw models status
openclaw gateway restart