What's New
✨ AI Agent Integration (Beta)
- Connect your AI assistant to your network data — Claude Desktop, Claude Code, and Gemini CLI can now query your UniFi Log Insight instance directly
- Ask your AI to search firewall logs, look up threat intelligence, review blocked traffic, review your firewall policy and suggest improvements, explore your network topology, and check system health — all through natural conversation
- Secure by design — every connection requires a token with granular permission scopes (logs, firewall, network, system, threats), and all AI activity is logged in an audit trail
- Built on the open Model Context Protocol (MCP) standard — works with any compatible desktop client
Easy Setup
- New Settings → MCP panel to enable the server, manage tokens, and view AI activity logs
- One-click config snippets — pre-filled connection configs for Claude Desktop and CLI commands for Claude Code and Gemini CLI, ready to copy and paste
- Token permissions let you control exactly what your AI can access
Dashboard Improvements
- Log ID shown in expanded log rows — useful for referencing specific events when working with an AI agent
- Improved copy buttons across all log detail fields with consistent styling
- Filter dropdowns no longer hidden behind the table header when scrolling
- Light theme support for all new elements
Key Files
receiver/routes/mcp.py— MCP serverui/src/components/SettingsMCP.jsx— MCP settings panelreceiver/routes/threats.py— Threat lookup APIui/src/components/CopyButton.jsx— Shared copy button component