Show HN: Starbase – Browser-based MCP server testing with AI chat integration

2 weeks ago 1

Connect, explore, and test any MCP server with AI models

🚀 Try Live Demo • Quick Start • Features • Documentation • Contributing

MCP servers let AI models interact with tools and data sources. The problem is that testing and debugging them is difficult. When you're building an MCP server, you need to see if your tools actually work with real AI models. When you're evaluating existing servers, you need to understand what they do without diving into source code or setting up complex test environments.

Anthropic's MCP Inspector helps with basic inspection, but it doesn't let you actually test tools with AI models. You can see what's available, but you can't see how it behaves in practice.

Starbase solves this by letting you connect any remote MCP server and immediately test it with Claude, GPT-4, or other models. No setup, no boilerplate. You get a working environment to develop, test, and explore MCP servers.

📡 Universal MCP Connectivity - Connect to any remote MCP server with just a URL. No complex setup or configuration required.
🤖 Multi-Model Integration - Test your MCP servers with leading AI models including Claude, GPT-4, and more. See how tools perform in real-world scenarios.
🔋 Batteries Included - OAuth flows, client registration, connection management, and more are all handled automatically.
⚡ Interactive Playground - Explore available tools, prompts, and resources from any MCP server through an intuitive interface.
🔧 Developer Friendly - Perfect for developing and debugging MCP servers. Test your changes instantly without additional tooling.
🔒 Secure by Default - Safe connection handling with proper authentication and authorization flows.
🎯 Zero Account Required - Start testing immediately with our hosted version at starbase.sh.

Hosted Version (Recommended)

The fastest way to get started is to use our hosted version:

Visit starbase.sh
Enter your MCP server URL
Start exploring and testing immediately

No installation, no account creation, no hassle.

Want to run Starbase locally? Follow these steps:

Node.js 18+ or Bun
PostgreSQL database (or any Prisma-supported database)

# Clone the repository git clone https://github.com/metorial/starbase.git cd starbase # Install dependencies npm install # or bun install # Set up environment variables cp .env.example .env # Edit .env with your configuration # Set up the database npm run db:push # or bun run db:push # Start the development server npm run dev # or bun run dev

Visit http://localhost:3000 to see Starbase in action!

starbase/ ├── app/ # Next.js app directory (pages, layouts, API routes) ├── components/ # React components ├── contexts/ # React context providers ├── lib/ # Utility functions and core logic ├── prisma/ # Database schema and migrations └── types/ # TypeScript type definitions

Framework: Next.js with App Router
UI: React with Styled Components
AI SDKs: Anthropic SDK, OpenAI SDK, Vercel AI SDK
MCP: Official Model Context Protocol SDK
Database: Prisma ORM with PostgreSQL
Authentication: NextAuth.js v5
Language: TypeScript

Create a .env file based on .env.example and configure:

# Database DATABASE_URL="postgres://..." # Authentication (optional for local development) AUTH_SECRET="your-secret-key" # AI Model API Keys (optional - add only the models you want to use) ANTHROPIC_API_KEY="your-anthropic-key" OPENAI_API_KEY="your-openai-key" # Other optional configurations NEXT_PUBLIC_BASE_URL="http://localhost:3000"

npm run dev - Start development server
npm run build - Build for production
npm run start - Start production server
npm run lint - Run ESLint
npm run db:generate - Generate Prisma client
npm run db:push - Push schema changes to database
npm run db:migrate - Run database migrations
npm run db:studio - Open Prisma Studio

We welcome contributions from the community! Whether you're fixing bugs, adding features, improving documentation, or reporting issues, your help makes Starbase better for everyone.