AI导哈

API文档MCP服务器

Duckfly MCP 是一个本地 Model Context Protocol 服务器，将 Duckfly 文档化的 API 以工具、提示词和资源的形式开放给 AI 客户端，并将工具调用转发到你的后端。

  • by duckfly-dev
  • 本地部署
  • 1 星标
  • 更新于
在 GitHub 查看 duckfly-dev/mcp

使用说明

Duckfly MCP

Run a local MCP server from your Duckfly API spec. Tools, prompts, and resources are generated from your documentation; tool calls are forwarded to your API.

Overview

Duckfly MCP is a local MCP (Model Context Protocol) server that uses the spec produced by Duckfly Core for your documented API. It exposes tools, prompts, and resources to AI clients and forwards tool calls to your backend.

Use the same Duckfly token as the proxy plugin. The spec is fetched once at startup from Duckfly Core.

Why use Duckfly MCP

  • Expose your API as MCP tools so AI assistants can call your endpoints
  • No manual MCP spec: tools and schemas come from your Duckfly documentation
  • Optional auth: None, Basic, Bearer, or OAuth (JWT via JWKS)
  • JSON-RPC over HTTP with optional SSE (single global stream or streamable per-request responses)

Quick Start

Global install

npm install -g @duckfly/mcp

Or run with npx

npx @duckfly/mcp

The CLI will ask for:

Prompt Description Default
Token Your Duckfly application token (same as proxy) —
MCP Port Port where the MCP server listens 9090
Target URL Base URL of your API (tool calls are sent here) http://localhost:3000
Auth mode None / Basic / Bearer / OAuth none

Configuration is saved in .duckfly-mcp.json.

How it works

┌─────────────┐           Duckfly Spec         ┌──────────────┐
│ Duckfly MCP │ <───────────────────────────── │ Duckfly Core │
│ (local)     │                                │ (your spec)  │
└──────┬──────┘                                └──────────────┘
       │
       │ Tool calls (e.g. POST /api/sample/123)
       ▼
┌──────────────┐
│ Your API     │  (Target URL)
└──────────────┘
  1. At startup, the plugin fetches the MCP spec from Duckfly Core (tools, prompts, resources).
  2. AI clients connect to your local MCP server and see those tools.
  3. When a client invokes a tool, the server calls Target URL + path (e.g. http://localhost:3001 + /api/sample/123).
  4. The response is returned to the client; the body includes a requested field (method + URL) for debugging.

If you get 404 on a tool call: the target server does not have that route. Ensure the service at Target URL is the same API documented in Duckfly, and if your API uses a prefix (e.g. /v1), set Target URL to http://localhost:3001/v1.

Endpoints

Method Path Description
GET /mcp Open the global SSE channel (one at a time).
POST /mcp JSON-RPC. With Accept: text/event-stream you get a streamed response; otherwise, if the global SSE is open, the response is sent there and the POST gets 202; else you get application/json.
DELETE /mcp End the session (requires Mcp-Session-Id).
GET /.well-known/oauth-protected-resource/mcp OAuth protected-resource metadata (only when OAuth is enabled).
GET /health Health check ({ "ok": true }).

License

MIT License

This license applies only to the Duckfly MCP package. The Duckfly platform and services are governed by separate terms.

Made with 🦆 by Duckfly

相关 MCP