Limitless AI MCP Server

What is Limitless AI MCP Server?

Limitless AI MCP Server is an advanced Model Context Protocol (MCP) server that enables AI assistants to interact with Limitless AI Pendant recordings. It provides a seamless bridge between AI tools and the Limitless API, allowing users to extract insights, search through recordings, and analyze their Pendant data.

How to use Limitless AI MCP Server?

First, obtain a Limitless API key from limitless.ai/developers. Then install either globally from npm (npm install -g limitless-ai-mcp-server) or from source. Configure your MCP client (e.g., Claude Desktop or Claude Code CLI) with the server path and API key environment variable. After configuration, you can chat naturally: “Show me my recordings from today.”

Key features of Limitless AI MCP Server

• Full Limitless API integration for Pendant recordings
• Advanced keyword search across recordings
• List recordings by date, date range, or most recent
• Five MCP tools for searching and retrieving lifelogs
• Pre-built prompt templates for common analysis tasks
• Intelligent caching with LRU eviction and configurable TTL

Use cases of Limitless AI MCP Server

• Search past conversations for specific topics or keywords
• Summarize recordings from a particular day or date range
• Extract action items and key decisions from meetings
• Identify recurring themes or patterns in personal recordings
• Generate structured meeting notes from raw lifelog content

FAQ from Limitless AI MCP Server

What are the system requirements?

Node.js 22+ is required. You also need a Limitless Pendant (the API only returns Pendant recordings) and a Limitless API key from limitless.ai/developers.

How do I configure the server for Claude Desktop?

Open your Claude Desktop config file (location varies by OS) and add a mcpServers entry with the command node, the path to dist/index.js, and the LIMITLESS_API_KEY environment variable. Restart Claude Desktop for changes to take effect.

What tools does the server provide?

Five tools: limitless_get_lifelog_by_id, limitless_list_lifelogs_by_date, limitless_list_lifelogs_by_range, limitless_list_recent_lifelogs, and limitless_search_lifelogs. Each accepts parameters for filtering, sorting, and content inclusion.

How does caching work?

The server uses an LRU (Least Recently Used) eviction strategy with configurable TTL (Time To Live). Separate caches exist for lifelogs (default size: 100, TTL: 5 minutes) and search results (default size: 50, TTL: 3 minutes). Environment variables allow customization.

Is this server production ready?

This is a beta release (v0.0.1). Core features are implemented and tested, but community feedback is needed to validate all functionality. Users should report issues on GitHub.