C# MCP Server Extension

This VS Code extension provides a Model Context Protocol (MCP) server that exposes C# language features to AI tools. It enables advanced code navigation and analysis capabilities for C# code when using AI coding assistants that support the MCP protocol.

Features

• Find Usages: Locate all references to a symbol in your C# codebase with detailed context
• HTTP/SSE Server: Exposes language features over an MCP-compatible HTTP server
• AI Assistant Integration: Ready to work with AI assistants that support the MCP protocol

Requirements

• C# for Visual Studio Code extension must be installed
• .NET SDK (compatible with the C# extension)

Installation

1. Install the C# for Visual Studio Code extension
2. Install this extension
3. Open a C# project in VS Code

Usage

The extension will automatically start an MCP server when activated. To configure an AI assistant to use this server:

1. The server runs on port 8008 by default
2. Configure your MCP-compatible AI assistant to connect to:
   • SSE endpoint: http://localhost:8008/sse
   • Message endpoint: http://localhost:8008/message

Available Commands

• C# MCP: Start Server - Manually start the MCP server
• C# MCP: Stop Server - Stop the running MCP server

Tools

Find Usages

Find all references to a C# symbol in your codebase:

{
  "name": "find_usages",
  "arguments": {
    "textDocument": {
      "uri": "file:///path/to/your/file.cs"
    },
    "position": {
      "line": 10,
      "character": 15
    },
    "context": {
      "includeDeclaration": true
    }
  }
}

The tool returns:

• File paths for each reference
• Line and character positions
• Code preview for each reference

TODO

Endpoint	Purpose for MCP Server (LLM Assistant)
/codecheck	Runs full code diagnostics (errors, warnings, linting) and returns issues. LLM can use this to detect and fix issues.
/codeformat	Formats code properly. Useful for cleaning up LLM-generated code before inserting it into the project.
/completion	Provides autocomplete suggestions. The LLM can use this to enhance its own completions by comparing them with the language server’s.
/completion/resolve	Further resolves completion items, adding details like signatures and documentation.
/findsymbols	Finds all declared symbols (functions, classes, variables), helping the LLM understand the codebase.
/findusages	Finds where a function, class, or variable is used. Useful for refactoring suggestions.
/gototypedefinition	Finds the type definition of a symbol (e.g., where a class is defined). Helps the LLM provide type-aware completions.
/findimplementations	Retrieves implementations of an interface or base class. Helps the LLM when suggesting method overrides.
/getcodeactions	Lists available refactoring or quick fixes. The LLM can suggest applying these fixes.
/runcodeaction	Executes a refactoring or quick fix. If the LLM recommends a fix, it can trigger this action automatically.
/rename	Renames a symbol across the project. If the LLM suggests a rename, it can apply this action.
/typelookup	Retrieves type information for a symbol. Useful for an LLM to provide type-aware autocompletions.
/metadata	Retrieves project metadata (e.g., dependencies, imports). Helps the LLM understand frameworks and libraries in use.
/runfixall	Applies all available quick fixes. If the LLM detects multiple issues, it can run this to fix everything in one step.
/quickinfo	Retrieves hover info (docs, type hints, signatures). Useful for explaining code in chat-based LLM assistants.

Troubleshooting

If you encounter issues:

1. Ensure the C# extension is properly installed and activated
2. Check that the C# project has loaded correctly
3. Verify that port 8003 is available on your system

Contributing

Please feel free to submit issues or pull requests to the GitHub repository.

License

This extension is licensed under the MIT License.