Disconnected No Reason (1006)

[left wyvern]

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

[arctic thicket]

Hey! 🦐 Thanks for the detailed report. That 1006 disconnect is definitely odd — 45 seconds is a long time to auto-recover.

First: What version are you running? (clawdbot --version) — version mismatches are the #1 cause of weird bugs.

Quick diagnostics to run:

# Check gateway status
clawdbot daemon status

# Check if MiniMax is actually reachable
clawdbot status --deep

# Watch logs for the disconnect
clawdbot logs --follow

Common culprits for 1006 on fresh Linux installs:

1. MiniMax API connectivity — MiniMax servers can be flaky. Try switching to a different model temporarily to isolate:

   clawdbot model list  # see available
   

2. Telegram connection issues — The control console disconnects when the gateway loses its WebSocket link. Check:

   clawdbot channels status --probe
   

3. Systemd/daemon issues — On fresh Linux Mint, the daemon might not be fully installed:

   clawdbot daemon install
   clawdbot daemon restart
   

4. Firewall/network — Check if something's interfering:

   sudo ufw status  # or whatever firewall you have
   

What to look for in logs:
Run clawdbot logs --follow and trigger a message — the logs will show exactly why it's disconnecting.

Let me know what clawdbot --version says and what the logs show, and we can narrow it down! 🎯

[nocturne spindle]

Clawdbot 2026.1.24-3 (885167d)

[arctic thicket]

Thanks! Version is recent (2026.1.24-3).

Let's dig deeper:

# Run deep diagnostics
clawdbot status --deep

# Check recent logs for the actual error
clawdbot logs --limit 100 | grep -i "error\|disconnect\|mini\|telegram"

# Check gateway health
clawdbot health --json

Quick things to try:

1. Test MiniMax directly — temporarily switch to a different model to rule out MiniMax API issues:

   # Check what models you have
   clawdbot model list
   
   # Switch temporarily (e.g., to o4-mini or similar)
   clawdbot model set <model-name>
   

2. Check if it's Telegram or MiniMax — disable one at a time in your config:

   # In your ~/.clawdbot/clawdbot.json, set enabled: false for channels.telegram
   # Then restart: clawdbot daemon restart
   

3. Resource check — fresh Linux install might have limits:

   free -h           # memory
   ulimit -a         # open files
   systemctl status clawdbot-gateway  # or check journalctl

The 45-second auto-recovery is the key clue — it suggests something external (API timeout?) is causing the gateway to restart or reconnect.

What do clawdbot logs --follow show when the disconnect happens? That's where the real error will be.

Found it! The gateway is unreachable and the service shows state failed — that's why you're disconnecting. The gateway process keeps dying and auto-restarting (hence the 45s recovery).

Check why the service is failing:

# See the actual error
journalctl --user -u clawdbot-gateway.service -n 50 --no-pager

# Or check file logs
tail -50 /tmp/clawdbot/clawdbot-*.log

Quick fixes to try:

# 1. Restart the service manually to see errors

clawdbot daemon stop
clawdbot gateway --verbose  # Run foreground to see errors in real-time

# 2. If port conflict, kill the existing process
lsof -nP -iTCP:18789 -sTCP:LISTEN

# 3. Reinstall the service
clawdbot daemon uninstall
clawdbot daemon install
clawdbot daemon restart

Most common causes on fresh Linux Mint:

• Something else already on port 18789
• Permission issues with ~/.clawdbot/
• Node version mismatch (should be Node 22+)

Run journalctl --user -u clawdbot-gateway.service -n 50 --no-pager and paste what it says — that'll tell us exactly why it's failing! 🔍