Code-Graph-RAG is an open-source tool designed for AI-powered analysis and understanding of codebases across multiple programming languages. It builds comprehensive knowledge graphs from code using Tree-sitter-based parsing, enabling users to visualize and explore code structures such as functions, classes, modules, imports, and call relationships. The platform is aimed at developers and teams seeking advanced codebase intelligence, optimization, and compliance mapping.
A key feature of Code-Graph-RAG is its ability to parse code in 12 languages—including Python, TypeScript, JavaScript, Rust, Java, C, C++, Lua, C, Go, PHP, and Scala—using a unified graph schema for consistent querying. The tool stores these interconnected graphs in Memgraph and allows users to query, edit, and optimize code using natural language. AI translates plain English questions into precise Cypher graph queries, enabling tasks such as identifying authentication functions or searching for specific code patterns. The platform also supports AI-driven code editing, with AST-based targeting for function modifications, visual diff previews, and exact code block replacements. Optimization workflows are enhanced by language-specific best practices and interactive approval processes.
Code-Graph-RAG integrates with Claude Code as an MCP server, providing ten tools for querying, editing, searching, and optimizing codebases directly from the IDE. These tools cover tasks such as retrieving function code, replacing code, exporting graph data, semantic search, and executing shell commands. The system also allows for the definition of custom knowledge graph relationships, supporting advanced use cases like compliance traceability, data lineage, security classification, team ownership, and cost attribution.
Deployment options include fully managed cloud-hosted solutions, on-premise installations, and air-gapped environments for regulated industries. The open-source community edition is free to use, while enterprise offerings provide dedicated support, consulting, custom development, technical support contracts, integration consulting, and training services. Code-Graph-RAG positions itself as a comprehensive toolkit for codebase analysis, intelligence, and optimization, with extensibility for organizational and compliance needs.
Code-Graph-RAG sits in PulseGate's AI category. It focuses on understanding and optimizing large, multi-language codebases using AI-powered knowledge graphs and natural language queries. Code-Graph-RAG is an open-source project aimed at software developers and engineering teams managing complex codebases. The project is open source (MIT). It runs on the command line, and it can be self-hosted.
Code-Graph-RAG first shipped in 2025. Development happens publicly on GitHub with 2.3k stars and 251 commits in the last 90 days. Among its 6 catalogued features are knowledge graph generation, multi-language parsing, and natural language queries. It exposes integrations via an MCP server.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do