Code2Flow MCP 服务器

What is Code2Flow MCP 服务器?

Code2Flow MCP 服务器 wraps the code2flow command-line tool into an MCP (Model Context Protocol) server. It allows AI applications to generate and access code call graphs through the standardized MCP protocol. The server is intended for developers who need automated call graph generation from source code across multiple languages.

How to use Code2Flow MCP 服务器?

Install Python 3.7+, clone the repository, create a virtual environment, install dependencies (pip install -r requirements.txt), and install the code2flow CLI tool. Start the server with python server.py or use MCP tools like mcp dev server.py. You can also integrate it into Claude Desktop via mcp install server.py or into Cursor via a JSON configuration. The server exposes MCP tools such as generate_call_graph, check_code2flow_version, and analyze_code_complexity.

Key features of Code2Flow MCP 服务器

• Analyzes source code and generates call graphs.
• Supports Python, JavaScript, Ruby, and PHP.
• Serves call graphs over the MCP protocol for AI integration.
• Outputs call‑graph images in PNG format.
• Provides version checking of code2flow.
• Offers code complexity analysis.

Use cases of Code2Flow MCP 服务器

• AI assistants automatically analyze code dependencies from user projects.
• Developers visualize call relationships in legacy or large codebases.
• Automated documentation generation by extracting call graphs.
• Code review tools assess complexity and call depth.

FAQ from Code2Flow MCP 服务器

What programming languages does Code2Flow MCP 服务器 support?

It supports Python, JavaScript, Ruby, and PHP. The language parameter in the generate_call_graph tool accepts python, js, ruby, and php.

What are the runtime requirements to run Code2Flow MCP 服务器?

You need Python 3.7 or later, the code2flow package installed via pip install code2flow, and the dependencies listed in requirements.txt. The server runs on Windows 11 and other operating systems with PowerShell or a command‑line terminal.

How do I generate a call graph using the server?

Call the generate_call_graph tool with the required source_paths parameter (list of files or directories). Optionally set output_path, language, exclude, or include patterns. The server returns an image resource ID that can be accessed via the call-graph:// resource URI.

What MCP resources are available on Code2Flow MCP 服务器?

Three resources are provided: help://code2flow for documentation, languages://supported for the list of supported languages, and call-graph://<imageID> for the generated call‑graph images.

Can I use Code2Flow MCP 服务器 with Claude Desktop or Cursor?

Yes. Use mcp install server.py to add it to Claude Desktop. For Cursor, add a JSON entry to the MCP configuration specifying the command cmd, argument /c, and the Python command to run server.py.