mcp-server-tree-sitter
The MCP Tree-sitter Server is a Model Context Protocol server designed for code analysis, enabling AI assistants to intelligently interact with various coding languages through structured access and context management functionalities.
MCP Tree-sitter Server
A Model Context Protocol (MCP) server that provides code analysis capabilities using tree-sitter, designed to offer AI assistants intelligent access to codebases with appropriate context management. It is language-agnostic, supports a variety of programming languages, and is structure-aware using AST-based understanding.
Features
- Flexible Exploration: Examine code at multiple levels of granularity
- Context Management: Provides information without overwhelming context
- Language Agnostic: Supports Python, JavaScript, TypeScript, Go, Rust, C, C++, Swift, Java, Kotlin, Julia, and APL
- Searchable: Find specific patterns using search and queries
- Caching: Optimized performance through parse tree caching
- Symbol Extraction: Analyze functions, classes, and code symbols
- Dependency Analysis: Identify code dependencies
- State Persistence: Maintains project registration and cached data
- Secure: Built-in security boundaries and input validation
Installation
See the repository for installation instructions using Python pip and development setups.
Usage
Register projects, explore files, analyze code structure, search and analyze code complexity directly within the MCP server or using Python code.