Codebase Analyzer is a comprehensive tool designed for in-depth analysis of Python codebases. It leverages Abstract Syntax Trees (AST) to parse code, extract key elements like functions, classes, and variables, and store this information in a Weaviate vector database. By utilizing Google's Gemini models for optional LLM enrichment, it generates semantic descriptions and embeddings, enabling Retrieval-Augmented Generation (RAG) for answering natural language questions about the code. The tool exposes analysis and querying capabilities through a Model Context Protocol (MCP) server, offering features such as file watching, codebase dependency management, cross-codebase querying, and MermaidJS call graph visualization, providing a powerful solution for understanding and navigating complex Python projects.
Características Principales
01Answers natural language questions about the codebase using RAG.
02Stores code elements and vector embeddings in Weaviate.
03Optionally generates semantic descriptions and embeddings using Gemini.
04Parses Python code using AST to extract code elements and relationships.
053 GitHub stars
06Exposes analysis and querying capabilities through an MCP server.