What are the runtime requirements?
You need Node.js (for `npx`) and `uv` installed. The server also requires ProxyPin to be installed, open, and configured to cache history (enable “Cache Date”).
@ElonJask
MCP Server for ProxyPin - expose HTTP capture data to AI tools
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"proxypin": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"proxypin-mcp@latest"
]
}
}
}工具列表暂未提供
当前收录暂时没有可展示的工具列表,不代表服务没有工具。请查阅项目文档了解支持的能力。
点击上方按钮,在临时沙箱中运行此服务器并实时获取工具列表。
Proxypin Mcp Server is a Model Context Protocol (MCP) server that integrates with the open‑source ProxyPin tool to expose your local HTTP capture history to MCP‑capable clients. It reads ProxyPin’s HAR files and provides tools for listing, searching, and analyzing captured requests, as well as generating API code snippets.
Install and open ProxyPin, enable “Cache Date” in History settings, and ensure some capture history exists. Run npx -y proxypin-mcp@latest (the NPX bridge tries uvx first then falls back to uv tool run). Add the following configuration to your MCP client’s config file:
{
"mcpServers": {
"proxypin": {
"type": "stdio",
"command": "npx",
"args": ["-y", "proxypin-mcp@latest"]
}
}
}
If your client does not support type, remove that field.
You need Node.js (for `npx`) and `uv` installed. The server also requires ProxyPin to be installed, open, and configured to cache history (enable “Cache Date”).
It reads HAR files from ProxyPin’s local data directory. The path can be set via the `PROXYPIN_DATA_DIR` environment variable or auto‑detected. No network connection to ProxyPin is needed.
The server exposes recent requests (with optional filters), single request details including body, search results across URL and bodies, a list of captured domains with request counts, and API structure analysis per domain.
It works with Windsurf, Cursor, Claude Desktop, and Codex. Any client that can run a stdio MCP server should be compatible.
By default, the server scans up to 50 HAR files (`PROXYPIN_HAR_LIMIT`) and keeps at most 102,400 bytes of body content per request (`PROXYPIN_MAX_BODY_SIZE`). These can be adjusted via environment variables.
Painless E2E Automation for Mobile and Web

DeepMark helps teachers deliver rapid, consistent marking with meaningful feedback for every student — in a fraction of the time. What once took a week, now takes one free period.
AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
A curated list of references for MLOps
Hindsight: Agent Memory That Learns
评论