codegraph-rust
100% Rust implementation of code graphRAG with blazing fast AST+FastML parsing, surrealDB backend and advanced agentic code analysis tools through MCP for effic
What is codegraph-rust?
100% Rust implementation of code graphRAG with blazing fast AST+FastML parsing, surrealDB backend and advanced agentic code analysis tools through MCP for efficient code agent context management
Overview of codegraph-rust
What it is
CodeGraph is an open-source tool that transforms a codebase into a semantically searchable knowledge graph. It is implemented in 100% Rust and uses AST (Abstract Syntax Tree) parsing, FastML parsing, a SurrealDB backend, and MCP (Model Context Protocol) for agentic code analysis. The tool is designed to give AI coding assistants persistent, structured understanding of a project's architecture, relationships, and documentation, rather than relying on simple text search or grep.
Key features
- Builds a real knowledge graph from code, preserving relationships such as function calls, dependencies, and architectural context.
- Supports multiple indexing tiers (
fast,balanced,full) to trade off between indexing speed and graph richness. - Enables hybrid search combining graph traversal and semantic embeddings.
- Provides agentic tools via MCP and CLI, including project-local hooks for Claude Code/Codex.
- Includes incremental indexing with caching of unchanged sources and deferred embedding/semantic resolution policies.
- Supports multiple languages (Rust, TypeScript/JavaScript, Python, Go, Java, C/C++) with LSP-based symbol resolution in higher tiers.
Use cases
- Providing AI coding assistants with long-term project memory across sessions.
- Impact analysis: understanding what code depends on a given function or module.
- Architecture discovery: surfacing package cycles, module relationships, and dataflow edges.
- Documentation linking: connecting symbols in README and docs to their definitions in code.
Getting started
bashcodegraph index /path/to/project --languages Rust --index-tier balancedFor full setup instructions, see the repository README.
License and pricing
The license is not stated in the provided metadata or README. The project is listed as open-source with no pricing information.
Based on the project's public README and repository metadata.