OpenClaw IRC

An IRC interface for OpenClaw, enabling access to Claude AI through standard IRC clients.

Why IRC?

• Minimal attack surface - No browser, no JavaScript supply chain, no WebSocket complexity
• Mature protocol - Decades of stability, well-understood security model
• Universal client support - Works with any IRC client on any platform
• Simple access control - Firewall at Layer 3, authenticate via VPN if needed

This project is part of a broader philosophy of securing self-hosted AI services through simplicity rather than complexity. See the KISS Security for OpenClaw whitepaper for the full rationale.

Features

• SASL PLAIN authentication over TLS (recommended)
• Auto-join configured channels
• Send/receive messages (channels and DMs)
• Mention detection (@botname)
• Message chunking for IRC's message length limits
• Reconnection with exponential backoff
• Message sanitization (strips IRC formatting/control characters)
• CTCP ACTION support (/me messages)
• Multi-account support
• Comprehensive test suite

Requirements

• OpenClaw instance
• IRC server (public like Libera.Chat, or self-hosted)
• Anthropic API key configured in OpenClaw

Installation

From Local Path

# Copy to OpenClaw extensions directory
cp -r . ~/.openclaw/extensions/irc
cd ~/.openclaw/extensions/irc
npm install --production

From npm (after publishing)

openclaw plugins install @openclaw/irc

Configuration

Add to ~/.openclaw/openclaw.json:

{
  "channels": {
    "irc": {
      "enabled": true,
      "server": "irc.libera.chat",
      "port": 6697,
      "ssl": true,
      "nickname": "openclaw",
      "username": "openclaw",
      "realname": "OpenClaw IRC Agent",
      "sasl": {
        "username": "openclaw",
        "password": "your-password"
      },
      "channels": ["#your-channel"],
      "dm": {
        "policy": "pairing",
        "allowFrom": []
      },
      "groupPolicy": "allowlist",
      "groups": {
        "#your-channel": {
          "users": ["*"]
        }
      }
    }
  }
}

See docs/installation.md for detailed configuration options.

Security Model

This plugin is designed to be deployed behind network-layer security controls:

1. Firewall inbound - Permit only authorized source IPs
2. VPN for remote access - Authenticate at the network layer with MFA
3. Egress filtering - Restrict outbound to required destinations only
4. No exposed authentication endpoints - The service is simply unreachable to attackers

For detailed security guidance, see KISS Security for OpenClaw.

Documentation

• Installation Guide
• Security Recommendations
• KISS Security Whitepaper

Tools

The tools/ directory contains helper scripts:

• openclaw-tunnel.sh - SSH tunnel launcher for secure remote access
• openclaw.desktop - Linux desktop entry for PWA-style access
• openclaw-icon.png - Application icon

Development

# Install dependencies
npm install

# Build
npm run build

# Run tests
npm test

# Run tests with coverage
npm run test:coverage

Contributing

Contributions welcome. Please read CONTRIBUTING.md first.

License

MIT License - see LICENSE

Star History

Links

• OpenClaw Project
• Project Documentation