Work90210/APIFold

Overview

APIFold generates a Model Context Protocol (MCP) server from any REST API using an OpenAPI specification. It requires no coding, making it a drop-in bridge between existing REST endpoints and MCP-compatible AI agents.

Best for

Best for
Developers who need a fast, no‑code way to make REST APIs available as MCP servers.

Use cases

• Convert internal REST APIs into MCP servers for AI assistants
• Rapidly expose third-party APIs to AI agents via MCP without writing server code
• Test MCP integration for existing REST endpoints before building custom servers

How to use

Install

npx apifold serve ./your-openapi-spec.yaml --base-url https://api.example.com

Tools exposed

• npx
• npm

Tested with

Claude, Cursor, Copilot

Example client config

spec: ./openapi/stripe.yaml\nport: 3001\ntransport: sse\nbaseUrl: https://api.stripe.com\nauth:\n  type: bearer\n  token: ${STRIPE_API_KEY}\nfilters:\n  tags: [payments, customers]\n  methods: [get, post]

Notes

APIFold generates a Model Context Protocol (MCP) server from any REST API using an OpenAPI specification. It requires no coding, making it a drop-in bridge between existing REST endpoints and MCP-compatible AI agents.

6 stars on GitHub. Last updated 2026-05-25. Licensed AGPL-3.0.

Use cases

• Convert internal REST APIs into MCP servers for AI assistants
• Rapidly expose third-party APIs to AI agents via MCP without writing server code
• Test MCP integration for existing REST endpoints before building custom servers

Pros

• Completely code‑free setup from an OpenAPI spec
• Leverages existing REST APIs without modification
• Quick to prototype MCP server integrations

Cons

• Requires a valid OpenAPI specification for the target API
• Low GitHub star count (6) indicates early‑stage or limited community adoption
• May not handle non‑standard or undocumented REST endpoints

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.