A personal fork of DeusData/codebase-memory-mcp — a knowledge graph-based code intelligence engine for AI coding assistants.

All credit for the core project goes to the original authors at DeusData. This fork adds personal improvements on top of their excellent work.

Indexes your codebase into a persistent knowledge graph using tree-sitter AST analysis across 66 languages. Exposes 14 MCP tools so AI agents (Claude Code, Codex CLI, Gemini CLI, etc.) can answer structural questions about your code in under 1ms — without reading every file.

Single static binary. Zero dependencies. SQLite-backed.

macOS / Linux:

curl -fsSL https://raw.githubusercontent.com/DeusData/codebase-memory-mcp/main/install.sh | bash

Windows (PowerShell):

Invoke-WebRequest -Uri https://raw.githubusercontent.com/DeusData/codebase-memory-mcp/main/install.ps1 -OutFile install.ps1
.\install.ps1

Build from source:

git clone https://github.com/rashidul-rahul/codebase-memory-mcp.git
cd codebase-memory-mcp
make -f Makefile.cbm

The binary lands in build/c/codebase-memory-mcp.

After installing, add to your MCP config (~/.claude/.mcp.json or project .mcp.json):

{
  "mcpServers": {
    "codebase-memory-mcp": {
      "command": "/path/to/codebase-memory-mcp",
      "args": []
    }
  }
}

Restart your agent. Say "Index this project" — done.

Once indexed, your AI agent can use the MCP tools directly. Some examples:

"What calls ProcessOrder?"
"Show me the architecture of this project"
"What changed in the last commit and what does it affect?"
"Find dead code in this repo"

The agent translates your questions into graph queries automatically.

Built on top of codebase-memory-mcp by DeusData. Licensed under MIT.