STiFLeR7/memex

What it does

STiFLeR7/memex is an MCP (Model Context Protocol) server that creates a persistent, temporal knowledge graph of your development work. It automatically watches your git repositories and extracts information about modules, symbols, architectural decisions, and open problems using Graphiti and Neo4j.

The system builds a sophisticated understanding of your codebase structure and evolution over time, with each relationship in the knowledge graph carrying a validity window and confidence score that degrades based on recency. This allows AI coding agents to maintain accurate, up-to-date context about your project.

How to set up

Installation requires Python 3.7+ and involves cloning the repository from GitHub, installing dependencies via pip, configuring your git repositories for monitoring, and setting up Neo4j for the knowledge graph backend. The tool connects to any AI agent supporting the MCP protocol, making it compatible with Claude, ChatGPT, and other AI coding assistants.

Key features

• Automatic git repository monitoring and analysis
• Temporal knowledge graph construction with Graphiti + Neo4j
• Confidence scoring and validity windows for graph edges
• MCP server protocol for seamless AI agent integration
• Cross-platform support (macOS, Windows, Linux)
• Symbol, module, and decision tracking
• Open problem identification and tracking