MCP Server Information

Server Name: CodeMunch Pro
Repository URL: https://github.com/BigJai/codemunch-pro
Brief Description: Intelligent code indexing MCP server with Tree-sitter AST extraction, hybrid search (FTS5 + vector), call graphs, 10 languages, and O(1) byte-offset symbol retrieval (99% token savings for AI agents).

Basic Requirements

• Open Source: MIT License
• MCP Compliant: Implements MCP API specification via FastMCP
• Active Development: Published to PyPI March 2026, active development
• Docker Artifact: Dockerfile present in repo root (installs from PyPI)
• Documentation: README with install, config, and tool reference
• Security Contact: GitHub Issues at https://github.com/BigJai/codemunch-pro/issues

Submitter Checklist

• This server meets the basic requirements listed above
• I understand this will undergo automated and manual review.
• I have tested the MCP Server using task validate -- --name SERVER_NAME
• I have built the MCP Server using task build -- --tools SERVER_NAME

About This Server

CodeMunch Pro gives AI coding agents surgical, token-efficient access to large codebases.

Key features:

• 13 tools — index, search, outline, get symbols, traverse call graphs
• 10 languages — Python, JavaScript, TypeScript, Go, Rust, Java, C, C++, C#, Ruby
• O(1) symbol retrieval — byte-offset stored per symbol, direct seek, no full-file reads
• Hybrid search — FTS5 BM25 + 384-dim FastEmbed vectors merged via Reciprocal Rank Fusion
• Incremental indexing — SHA-256 file hashing, only changed files re-parsed
• Call graph traversal — AST-traced function call edges, depth-traversable
• Local SQLite storage — per-repo DBs in ~/.codemunch-pro/, no cloud dependency

Install:

pip install codemunch-pro

MCP Config:

{
  "mcpServers": {
    "codemunch-pro": {
      "command": "codemunch-pro"
    }
  }
}

PyPI: https://pypi.org/project/codemunch-pro/