tree-sitter-analyzer and mcp-server-tree-sitter
These are competitors offering overlapping Tree-sitter-based code analysis capabilities via MCP, with A emphasizing multi-language scalability and CLI flexibility while B provides a more focused MCP server implementation.
About tree-sitter-analyzer
aimasteracc/tree-sitter-analyzer
A scalable, multi-language code analysis framework based on Tree-sitter, usable both as a CLI tool and an MCP server.
Combines tree-sitter's syntax parsing with `fd` and `ripgrep` for evidence-based symbol discovery, enabling minimal-context retrieval that reduces AI prompt tokens by up to 95%. Implements the SMART workflow (Set-Map-Analyze-Retrieve-Trace) to surface relevant code regions without full repository indexing, making it practical for messy or legacy codebases. Exposes MCP protocol tools for Claude, Cursor, and other AI assistants while maintaining project boundary security.
About mcp-server-tree-sitter
wrale/mcp-server-tree-sitter
MCP Server for Tree-sitter
Exposes AST-based code analysis through MCP with support for 11+ languages via tree-sitter parsers, enabling AI assistants to efficiently explore codebases with granular control over context depth. Implements parse tree caching, symbol extraction, and dependency analysis while maintaining state persistence across invocations. Integrates directly with Claude Desktop through stdio transport and provides optional tree-sitter query support for pattern-based code searches.
Scores updated daily from GitHub, PyPI, and npm data. How scores work