#connection error.
1 messages Β· Page 1 of 1 (latest)
@mystic bone
I see the Discord connection error notification and a message from NoLimits. The message is just a mention of user @mystic bone.
What would you like me to do? I can:
- Check the channel status
- Try to read the message in that channel
- Investigate the connection error
- Something else?
Let me know what you need! π¦
@mystic bone my bot replies to me saying "Connection Error." on all messaging channels. Signal, Discord, TUI, Gateway. Things I've tried.
-I can access the gateway website.
-clawdbot status shows gateway up and running and healthy.
i've restarted the gateway.
-logs don't show anything abnormal
Still the same issue.
Since the docs are limited, let me provide systematic troubleshooting for this "Connection Error" on all channels:
Most likely causes:
- Gateway is up but plugins can't reach it - Check the gateway URL in your plugin configs matches what
clawdbot gateway statusreports - Plugins aren't actually running - Even if gateway status looks healthy, the individual channel plugins (Signal, Discord, TUI) need to be running separately
- Gateway URL mismatch - If you recently changed gateway ports/hosts, plugins might be pointing to the old address
Quick diagnostic steps:
# 1. Confirm gateway is really running and get the URL
clawdbot gateway status
# 2. Check what URL your plugins are configured to use
# (depends on your setup - check ~/.clawd/ or your config directory)
# 3. Test gateway connectivity from the plugin host
curl -v http://YOUR_GATEWAY_URL/health
Common fix: If you restarted the gateway and the IP/port changed, you'll need to update the gateway URL in each channel plugin's configuration.
Could you share:
- Output of
clawdbot gateway status - Your gateway URL (from the status output)
- How you're running the plugins (Docker, npm, etc.)?
This will help pinpoint where the disconnect is. π¦
Clawdbot 2026.1.24-3 (885167d) β Because Threads wasn't the answer either.
(node:78425) [DEP0040] DeprecationWarning: The punycode module is deprecated. Please use a userland alternative instead.
(Use node --trace-deprecation ... to show where the warning was created)
β
β
Service: systemd (enabled)
File logs: /tmp/clawdbot/clawdbot-2026-01-28.log
Command: /usr/bin/node /home/finnigan/.npm-global/lib/node_modules/clawdbot/dist/entry.js gateway --port 18789
Service file: ~/.config/systemd/user/clawdbot-gateway.service
Service env: CLAWDBOT_GATEWAY_PORT=18789
Config (cli): ~/.clawdbot/clawdbot.json
Config (service): ~/.clawdbot/clawdbot.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 (pid 78141, state active, sub running, last exit 0, reason 0)
RPC probe: ok
Listening: 127.0.0.1:18789
Troubles: run clawdbot status
Troubleshooting: https://docs.clawd.bot/troubleshooting
@mystic bone
@mystic bone
Clawdbot 2026.1.24-3 (885167d) β Because Threads wasn't the answer either.
(node:78425) [DEP0040] DeprecationWarning: The punycode module is deprecated. Please use a userland alternative instead.
(Use node --trace-deprecation ... to show where the warning was created)
β
β
Service: systemd (enabled)
File logs: /tmp/clawdbot/clawdbot-2026-01-28.log
Command: /usr/bin/node /home/finnigan/.npm-global/lib/node_modules/clawdbot/dist/entry.js gateway --port 18789
Service file: ~/.config/systemd/user/clawdbot-gateway.service
Service env: CLAWDBOT_GATEWAY_PORT=18789
Config (cli): ~/.clawdbot/clawdbot.json
Config (service): ~/.clawdbot/clawdbot.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 (pid 78141, state active, sub running, last exit 0, reason 0)
RPC probe: ok
Listening: 127.0.0.1:18789
Troubles: run clawdbot status
Troubleshooting: https://docs.clawd.bot/troubleshooting
Found the issue! π―
Your gateway is bound to loopback only (127.0.0.1), meaning it only accepts connections from the local machine. Your plugins are likely running on a different machine or container and can't reach it.
The fix: Change the gateway bind address to 0.0.0.0 (all interfaces):
# Option 1: Via environment variable
export CLAWDBOT_GATEWAY_BIND=0.0.0.0
clawdbot gateway restart
# Option 2: Or update your service file
# Edit ~/.config/systemd/user/clawdbot-gateway.service
# Add: Environment="CLAWDBOT_GATEWAY_BIND=0.0.0.0"
systemctl --user daemon-reload
systemctl --user restart clawdbot-gateway
# Option 3: Or set in config
# In ~/.clawdbot/clawdbot.json:
# { "gateway": { "bind": "0.0.0.0" } }
After changing: The gateway will listen on all network interfaces, and your remote plugins should be able to connect.
β οΈ Security note: This exposes your gateway on your network. If you're running on a shared network, consider firewall rules or using a reverse proxy with authentication.
Let me know if you need help updating the service file! π¦
@mystic bone the gateway and plugins are on the same machine, this shouldn't matter.
You're right - on the same machine, loopback should work fine.
If everything is local and gateway is healthy, the issue is likely:
- Plugins aren't actually running - The gateway is up, but the channel plugins (Signal, Discord, TUI) might be stopped
- Wrong gateway URL in plugin configs - Plugins might be pointing to a different port/address
- Plugin authentication failures - Could be failing to connect to the actual messaging services (Discord API, Signal proxy, etc.)
Quick check:
# Are any plugins running?
clawdbot status
# Or check for running node processes
ps aux | grep clawdbot
What to look for:
- If
clawdbot statusonly shows gateway and no plugins β they're not started - If plugins show as "stopped" or missing β that's your problem
Can you run clawdbot status and share the output? That should show which plugins are actually running vs. the gateway. π¦
@mystic bone they are all running fine. it seems like the gateway cannot talk to the model.
That's the root cause! If the gateway can't reach the model, everything fails.
Quick diagnostics:
# Check model config
clawdbot model list
# Check recent logs for model errors
grep -i "model\|openai\|anthropic\|error" /tmp/clawdbot/clawdbot-2026-01-28.log | tail -50
Common causes:
- Invalid API key - Double-check your
CLAWDBOT_MODEL_API_KEYor config - Network block - Can't reach the model provider (common with cloud providers blocked in certain regions)
- Wrong model name - Typo in the model identifier
- Rate limiting - Hit quota limits
Test your API key directly:
# Test OpenAI
curl -H "Authorization: Bearer YOUR_API_KEY" https://api.openai.com/v1/models
# Test Anthropic
curl -H "x-api-key: YOUR_API_KEY" https://api.anthropic.com/v1/messages
What provider/model are you using? And can you check the logs for any actual error messages when the bot tries to respond? π¦
curl: (7) Failed to connect to api.anthropic.com port 443 after 2 ms: Connection refused
"Connection refused" on port 443 = network/firewall issue, not a Clawdbot problem.
Your machine can't reach Anthropic's API at all. This is an environmental issue:
Quick checks:
# 1. Can you reach any HTTPS site?
curl -I https://google.com
# 2. DNS resolution working?
nslookup api.anthropic.com
ping -c 2 api.anthropic.com
# 3. Check firewall rules
sudo iptables -L -n | grep 443
# or
ufw status
Common causes:
- Firewall blocking outbound 443 - Check with your network admin or hosting provider
- Corporate/proxy network - Need to configure proxy settings
- Docker network isolation - If running in Docker, the container might be isolated
- ISP blocking - Some ISPs block certain API endpoints
If you're on a VPS or cloud host:
# Check if outgoing connections work at all
curl -I https://8.8.8.8
To configure a proxy (if needed):
export HTTPS_PROXY="http://proxy:port"
export HTTP_PROXY="http://proxy:port"
Ask your hosting provider or network administrator about outbound HTTPS access. This is outside Clawdbot's control. π¦
This bot was working fine with no major changes. I can ping anthropic, no ufw setup. Seems like anthropic is rejecting me. However, Claude Code seems to work fine. Is it my token being rejected?
Interesting - if Claude Code works but curl fails, it's likely an authentication or header issue, not a network block.
Anthropic API requires specific headers. Try:
# With proper headers (this is what Clawdbot uses)
curl -H "x-api-key: YOUR_ANTHROPIC_KEY" \
-H "anthropic-version: 2023-06-01" \
-H "content-type: application/json" \
https://api.anthropic.com/v1/messages \
-d '{"model":"claude-sonnet-4-20250514","max_tokens":10,"messages":[{"role":"user","content":"hi"}]}'
If that fails but Claude Code works:
- Claude Code might be using a different API key - Check which key Claude Code is using vs. your Clawdbot config
- Claude Code might have a proxy configured - Check env vars:
env | grep -i proxy
- Clawdbot config might have wrong key - Compare:
# Your Clawdbot model config cat ~/.clawdbot/clawdbot.json | grep -A5 model
Quick test: Is Claude Code using the same API key file or environment variable as Clawdbot? π¦
it appears Anthropic has revoked that oauth token. running claude setup-token no longer shows you the token to copy an paste.
That's it! OAuth token revoked = no API access.
To fix:
- Get a new API key: Go to https://console.anthropic.com/settings/keys
- Create a new API key (not OAuth - use an API key directly)
- Update Clawdbot config:
# Edit your config nano ~/.clawdbot/clawdbot.json # Set the new key: # { "model": { "apiKey": "sk-ant-api03-..." } } # Or set environment variable export ANTHROPIC_API_KEY="sk-ant-api03-..."
- Restart gateway:
clawdbot gateway restart
Note: Use an API key (starts with sk-ant-api03-) rather than OAuth. OAuth tokens can expire/revoke; API keys are more stable for bot use.
Let me know once you've added the new key and if it's working! π¦
Anthropic killed oauth tokens for use with moltbot. You should update the docs/setup wizard.