code-graph-rag MCP
An MCP server and CLI that build a knowledge graph of a monorepo and let an agent query, understand and edit multi-language code
- 5,352 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Reads and indexes the entire repository code
- Can propose code edits
- Sends code content to a language model
Install
- Works in
Manual install
uv tool install "code-graph-rag[treesitter-full,semantic]"Installs the tool with language parsing and semantic search. A Memgraph database is required.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Reads and indexes the entire repository code
- Can propose code edits
- Sends code content to a language model
This is third-party code. Review the repository files before installing.
What it does
code-graph-rag parses a monorepo with a syntax tree and builds a knowledge graph: files, modules, functions, classes and the relations between them. Over that graph the agent answers architecture questions, finds implementations and proposes edits across several languages at once. It works as an MCP server and as a CLI, with the graph stored in a Memgraph database. Many languages are supported through tree-sitter, with semantic search over the graph.
Who it is for. For developers and architects working with large multi-language monorepos.
Good fit when
- You need to understand a large repository's architecture through an agent
- You need to trace relations between modules and functions across the code
- You want to query code over a knowledge graph, not just text search
Not a fit when
- The project is small and a knowledge graph is overkill
- You cannot run a Memgraph database
- The model the RAG relies on is unavailable
Example request
Explain how the payment module and the task queue are connected in this monorepoMCP
- Transport
- stdio
- Authentication
- not required
Limitations
Requires a Memgraph database for the graph and a language model for queries. Installation goes through uv or pipx with the treesitter and semantic extras. Availability depends on the chosen model provider, so it is marked unknown.
How to disable. Remove the server from the MCP config, uninstall the tool with uv tool uninstall code-graph-rag and stop the Memgraph database.
FAQ
Why is Memgraph needed?
It stores the code knowledge graph the agent uses to trace relations and answer questions.
Which languages are supported?
Many languages via tree-sitter; the set installs through the treesitter-full and semantic extras.
Discussion
No questions or reviews yet.
Related
MCP serversAn MCP server that builds a code graph of functions, classes, calls and routes and answers the agent's structural queries locally
The /graphify skill builds a knowledge graph of a project's code, docs and PDFs so the agent queries the graph instead of grepping
Local code graph for agents: an MCP server returns entry points, symbol links and change impact in one call
An MCP server that builds a code knowledge graph: semantic search, call graph, HTTP route tracing and change impact analysis.
GraphMind
graphmind
Local MCP server that builds a code graph and remembers project decisions, so the agent stops re-reading the repo every session
Code graph and project memory in one Rust program: the agent gets excerpts with file and line citations
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| code-graph-rag (this card) | no | not required | 5,352 | |
| CodeGraph (Rust) | no | not required | 915 | a Rust code graph for agents |
| Tach | no | 2,868 | Tach checks Python module dependencies, public interfaces and import cycles against explicitly configured project boundaries. |
README in short
code-graph-rag is a RAG system for a monorepo: it builds a code knowledge graph and lets an agent query, understand and edit multi-language projects. It runs as an MCP server and a CLI, uses tree-sitter for parsing and a Memgraph database for the graph. It installs through uv or pipx with extras for language parsing and semantic search. MIT licensed.