api200
API 200 is an open source API gateway to simplify 3rd-party integrations. Import endpoints, set up caching, retries, and mocks. Access all services via one URL. Monitor logs, track errors, and get alerts on API incidents.
claude mcp add --transport stdio api-200-api200 npx api200-mcp@latest \ --env USER_KEY="your-api-key-here"
How to use
API 200 provides an MCP server named api200 that you can run via the npx package api200-mcp. This MCP server acts as a gateway and integration layer for managing third-party APIs, offering features such as authentication management, request caching, automatic retries, mock responses, response transformation, fallbacks, and custom headers. With the MCP setup, you can configure and interact with endpoints through your MCP client (e.g., Claude Desktop or other MCP clients) to manage, test, and monitor API integrations in a unified way. The included documentation and tooling facilitate importing endpoints from OpenAPI and Postman, as well as observing endpoint statistics and logs through the MCP interface.
To use it, you’ll typically provide your API key via environment variables (e.g., USER_KEY) when configuring the MCP tool in your client. The api200 MCP server will handle routing, authentication, caching, and retries per endpoint, while exposing your integrated endpoints in a consistent, observable manner. If you’re self-hosting, you can run the MCP via npx api200-mcp@latest and supply the required credentials to enable secure access to your configured APIs.
How to install
Prerequisites:
- Node.js and npm
- npx (comes with npm)
- Optional: Docker and Docker Compose if you prefer the self-hosted setup described in the project README
Installation (using the MCP from npm registry):
# Install/run the API200 MCP server via npx
npx api200-mcp@latest
If you want to configure the MCP with a specific key (as shown in the README):
# Example invoking with environment variable support (adjust to your shell)
export USER_KEY=your-api-key-here
npx api200-mcp@latest
Self-hosted setup (alternative, based on the project’s self-hosted instructions):
# Clone the repository and install dependencies
git clone https://github.com/API-200/api200-selfhosted
cd api200-selfhosted
npm i
# Run setup script (localhost with admin privileges)
sudo node setup.js
# Or for non-localhost:
node setup.js
# Start services
docker-compose up -d
Accessing the installation:
- Frontend: http://<your-hostname-or-ip>:3000
- API Handler: http://<your-hostname-or-ip>:8080
Additional notes
Tips and common considerations:
- Use a strong USER_KEY and keep it secret; treat it like a password for your MCP.
- If using the self-hosted route, ensure Docker, Docker Compose, and Node.js/npm are properly installed and that port exposure (3000, 8080) matches your environment.
- When running locally, note the potential need to modify hosts (e.g., 127.0.0.1 kong) for localhost setups as mentioned in the docs.
- The API200 MCP supports importing endpoints from OpenAPI and Postman; take advantage of this to quickly bootstrap your gateway configuration.
- Keep your MCP package up to date to benefit from security patches and feature improvements.
Related MCP Servers
klavis
Klavis AI (YC X25): MCP integration platforms that let AI agents use tools reliably at any scale
aci
ACI.dev is the open source tool-calling platform that hooks up 600+ tools into any agentic IDE or custom AI agent through direct function calling or a unified MCP server. The birthplace of VibeOps.
superglue
superglue (YC W25) builds integrations and tools from natural language. Get production-grade tools for long tail and enterprise systems.
n8n-workflow-builder
AI assistant integration for n8n workflow automation through Model Context Protocol (MCP). Connect Claude Desktop, ChatGPT, and other AI assistants to n8n for natural language workflow management.
local -gateway
Aggregate multiple MCP servers into a single endpoint with web UI, OAuth 2.1, and profile-based tool management
runi
See the truth about your APIs runi is an open-source API client that verifies what AI generates. When 46% of developers don't trust AI output (Stack Overflow 2025 Developer Survey), you need a second opinion. runi is that opinion.