MCP Playground
MCP Playground — a "Postman for MCP" that connects to any MCP server, auto-discovers tools, runs AI-powered diagnostic tests on every tool using Gemini, and lets you invoke tools through natural chat with a live executio
Core trick
Has Gemini generate arguments for each discovered MCP tool and runs them as an automated test pass over Streamable HTTP or SSE.
Description
MCP Playground — a "Postman for MCP" that connects to any MCP server, auto-discovers tools, runs AI-powered diagnostic tests on every tool using Gemini, and lets you invoke tools through natural chat with a live execution graph. Built solo in one day.
README excerpt
# MCP Playground MCP Playground is a "Postman for MCP" style web app. The goal is simple: connect to a real MCP server, discover its tools, test them automatically, and then use the same tools through a chat interface while watching the execution graph in real time. ## What It Does - Connects to remote MCP servers over Streamable HTTP or SSE - Discovers available MCP tools - Runs a `Test all tools` diagnostic flow with AI-generated arguments - Lets users invoke tools through natural language chat - Visualizes tool execution as a live graph and trace ## Tech Stack - Next.js 16 - React 19 - Tailwind CSS - Vercel AI SDK - `@ai-sdk/react` - `@ai-sdk/google` - `@ai-sdk/mcp` - Gemini 3.1 Pro via Vercel AI Gateway or direct Google access ## Run Locally 1. Copy `.env.example` to `.env.local` 2. Add `AI_GATEWAY_API_KEY` or `GOOGLE_GENERATIVE_AI_API_KEY` 3. Run `npm install` 4. Run `npm run dev` Open `http://localhost:3000`. ## Quick Review Flow 1. Connect to `https://gitmcp.io/docs` 2. Click `Test all tools` 3. Open the execution graph 4. Try a natural language tool request in chat ## License MIT.