Mcp Codebase Index
@MikeRecognex
17 MCP query tools for codebase navigation — functions, classes, imports, dependency graphs, change impact. Zero dependencies. 87% token reduction.
Overview
What is mcp-codebase-index?
A structural codebase indexer with an MCP server for AI-assisted development. It parses source files into structural metadata (functions, classes, imports, dependency graphs, call chains) and exposes 18 query tools via the Model Context Protocol, enabling Claude Code and other MCP clients to navigate codebases efficiently without reading entire files. Requires Python 3.11+ and has zero runtime dependencies.
How to use mcp-codebase-index?
Install via pip install "mcp-codebase-index[mcp]", then set the PROJECT_ROOT environment variable and run as a console script (mcp-codebase-index) or Python module (python -m mcp_codebase_index.server). Configure with Claude Code (.mcp.json) or OpenClaw (openclaw.json) by specifying the command and environment variable.
Key features of mcp-codebase-index
- Zero runtime dependencies (only Python 3.11+)
- Automatic incremental re-indexing using git diff
- Persistent disk cache for instant startup
- 18 query tools including dependency graph and call chains
- Supports Python (AST), TypeScript/JS, Go, Rust, C# (regex)
- Sub-millisecond query times even on codebases over 1M lines
Use cases of mcp-codebase-index
- AI-assisted code navigation without reading entire files
- Finding symbol definitions and usages across a codebase
- Change impact analysis and dependency tracking
- Exploring cross-file call chains and import relationships
