Nuanced MCP Server

A Model Context Protocol (MCP) server that provides call graph analysis capabilities to LLMs through the nuanced library.

Overview

This MCP server enables LLMs to understand code structure by accessing function call graphs through standardized tools and resources. It allows AI assistants to:

• Initialize call graphs for Python repos
• Explore function call relationships
• Analyze dependencies between functions
• Provide more contextually aware code assistance

API

Tools

• initialize_graph

• switch_repository

• list_repositories

• get_function_call_graph

• analyze_dependencies

• analyze_change_impact
  • Analyze the impact of changing a specific function
  • Inputs:
    • `file_path` (string)
    • `function_name` (string)

Resources

• graph://summary

• graph://repo/{repo_path}/summary

• graph://function/{file_path}/{function_name}
  • Get detailed information about a specific function
  • Parameters:
    • `file_path` (string) - Path to the file containing the function
    • `function_name` (string) - Name of the function to analyze

Prompts

• analyze_function

• impact_analysis

• analyze_dependencies_prompt
  • Create a prompt to analyze dependencies of a file or module
  • Parameters (at least one required):
    • `file_path` (string, optional) - Path to the file to analyze
    • `module_name` (string, optional) - Name of the module to analyze

Usage with Claude Desktop

Add this to your `claude_desktop_config.json`

UV

{
  "mcpServers": {
    "nuanced": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/nuanced-mcp",
        "run",
        "nuanced_mcp_server.py"
      ]
    }
  }
}