io.github.jscoobyced/jsc-typescript-ast-mcp

TypeScript AST MCP Server

A Model Context Protocol (MCP) server that provides TypeScript AST (Abstract Syntax Tree) analysis capabilities. This server allows you to find references for classes and methods, analyze component trees, and find dependencies in TypeScript code using the ts-morph library.

Features

• MCP Server: Implements the Model Context Protocol for integration with AI assistants
• TypeScript Analysis: Uses ts-morph for powerful TypeScript AST parsing and analysis
• Reference Finding: Find all references to a given class method in your codebase
• Dependency Finding: Use the find_dependency tool to list all files that import a specified library.
• Component Tree Analysis: Use the component_tree tool to analyze React component structures in your codebase.
• The tools are part of the MCP server and can be invoked via Claude Code or other MCP clients.

Prerequisites

• Node.js (version compatible with the project)
• Yarn package manager

Setup

1. Clone the repository

2. Install dependencies:

   yarn install
   

3. Copy the example configuration files:

   cp .env.example .env
   

4. Update .env with your project's tsconfig path:

   PROJECT_TSCONFIG_PATH=/path/to/your/project/tsconfig.json
   

How to Run

Build the project:

yarn build

Test the MCP server:

yarn test:mcp

This will start the MCP server that can be connected to by MCP-compatible clients.

Installing in Claude Code

To use this MCP server in Claude Code:

1. Build the server:

   yarn build
   

2. Configure Claude Code:

• Copy the .mcp.example.json to .mcp.json (or update your existing one with the content of the MCP configuration)
• Edit your .claude/settings.local.json and add the MCP section (or the subset needed if you already have some MCPs configured)

  “enableAllProjectMcpServers”: true,
  “enabledMcpjsonServers”: [
    “typescript-ast”
  ]

• Start Claude Code
• Type the /mcp command and check the MCP is connected

Development

Update the code:

1. Make changes to the TypeScript source files in src/
2. Build the project:

   yarn build
   

3. Test the changes:

   yarn test:mcp
   

Environment Variables

The server uses the following environment variable:

• PROJECT_TSCONFIG_PATH: Path to the tsconfig.json file to use for project analysis (defaults to 'tsconfig.json')

Example Configuration Files

• .env.example: Example environment variables file
• .mcp.example.json: Example MCP configuration file (if applicable)

Tool Descriptions

The server provides three tools:

1. find_reference

Find all references to a given class method in your codebase.

• Input:
  • filePath: Path to the file
  • className: Name of the class
  • methodName: Name of the method
• Output: List of file paths with line numbers where the method is referenced

2. find_dependency

List all files that import a specified library.

• Input:
  • dependencyName: Name of the dependency to search for
• Output: List of file paths that import the specified dependency

3. component_tree

Analyze React component structures in your codebase.

• Input:
  • entryFilePath: Path to the entry file
  • maxDepth: Maximum depth of the component tree (default: 3)
  • data-id (optional): Attribute name to capture into TreeNode.dataId (example: data-attribute-id)
• Output: JSON representation of the component tree structure, including optional element metadata such as props, dataId, and onClick

License

MIT