Code2Flow MCP 服务器
About
Code2Flow MCP 服务器 wraps the code2flow command-line tool into a Model Context Protocol (MCP) server, allowing AI applications to generate and access code call graphs through a standardized MCP interface. It supports Python, JavaScript, Ruby, and PHP source code, outputs call…
Details
- Author
- MCP-Mirror
- Downloads
- 241
- Categories
- Other
Jump to
- Analyzes source code and generates call graphs.
- Supports Python, JavaScript, Ruby, and PHP.
- Outputs call graphs as PNG images.
- Provides code2flow version checking.
- Includes code complexity analysis capability.
- Integrates seamlessly with AI applications via MCP.
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Code2Flow MCP 服务器Command (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install by cloning the repository, creating a virtual environment, and installing dependencies (including code2flow). Start the server directly with python server.py, or use the MCP development tools (mcp dev server.py or mcp install server.py). Configuration for Cursor can be added as a JSON entry specifying the command and path to server.py. Once running, clients can call the available MCP tools (e.g., generate_call_graph) with parameters such as source_paths, language, and optional output_path, exclude, and include patterns.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"code2flow mcp \u670d\u52a1\u5668": {
"code2flow": {
"command": "python",
"args": [
"server.py"
]
}
}
}
}McpServers
{
"code2flow": {
"command": "python",
"args": [
"server.py"
]
}
}Code2Flow MCP 服务器
这是一个将 code2flow 命令行工具包装为 MCP (Model Context Protocol) 服务器的项目。它允许 AI 应用程序通过标准化的 MCP 协议生成和访问代码调用图。
功能特点
- 分析源代码并生成调用图
- 支持多种编程语言（Python、JavaScript、Ruby、PHP）
- 通过 MCP 协议提供服务，易于与 AI 应用集成
- 图像以 PNG 格式输出
- 提供版本检查和代码复杂度分析功能
安装要求
- Python 3.7+
- Windows 11 或其他支持的操作系统
- PowerShell 或其他命令行终端
- 已安装 code2flow 命令行工具
安装步骤
1. 克隆此仓库
git clone https://github.com/kursk-ye/code2flow-mcp-server.git
cd code2flow-mcp-server
2. 创建并激活虚拟环境（推荐）
python -m venv venv
.\venv\Scripts\Activate.ps1 # 在 PowerShell 中
3. 安装依赖项
pip install -r requirements.txt
4. 安装 code2flow 命令行工具
pip install code2flow
使用方法
直接使用 Python 运行
1. 启动 MCP 服务器
python server.py
使用 MCP 工具运行
1. 使用 MCP Inspector 工具
mcp dev server.py
2. 安装到 Claude Desktop
mcp install server.py
3. 添加到 Cursor MCP 配置
"code2flow": {
"command": "cmd",
"args": [
"/c",
"python",
"path/to/server.py"
]
}
可用工具
服务器提供以下 MCP 工具：
1. generate_call_graph - 生成代码调用图
2. check_code2flow_version - 检查 code2flow 版本
3. analyze_code_complexity - 分析代码复杂度
可用资源
服务器提供以下 MCP 资源：
1. help://code2flow - 帮助文档
2. languages://supported - 支持的语言列表
3. call-graph://图像ID - 生成的调用图图像
示例（使用 Python MCP 客户端）
import asyncio
from mcp.client import MCPClient
async def main():
# 连接到服务器
client = MCPClient("http://localhost:8000")
# 创建新会话
session = await client.create_session()
# 调用工具生成调用图
result = await session.call_tool("generate_call_graph", {
"source_paths": ["path/to/your/code"],
"language": "python"
})
# 获取资源 ID
print(result)
# 检查code2flow版本
version_info = await session.call_tool("check_code2flow_version")
print(version_info)
# 分析代码复杂度
complexity = await session.call_tool("analyze_code_complexity", {
"source_path": "path/to/your/code",
"language": "python"
})
print(complexity)
if __name__ == "__main__":
asyncio.run(main())
配置选项
生成调用图时支持以下参数：
- source_paths：要分析的源代码文件或目录的路径列表
- output_path：（可选）输出文件的路径
- language：（可选）源代码语言（python、js、ruby、php）
- exclude：（可选）要排除的文件模式列表
- include：（可选）要包含的文件模式列表
文件结构
- server.py - 主服务器代码
- mcp_client_example.py - 客户端示例代码
- test_server.py - 测试代码
- requirements.txt - 依赖文件
许可证
MIT
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
