Smithery Registry OpenAPI Server
LocalAbout Smithery Registry OpenAPI Server
Smithery Registry OpenAPI Server provides an OpenAPI interface to the Smithery Registry API. It lets OpenAPI-compatible clients search MCP servers, retrieve their details, and generate WebSocket launch URLs with configuration.
Key features
- 01List and filter available MCP servers
- 02Semantic search and advanced query filters
- 03Retrieve server details and connection configuration
- 04Generate configured WebSocket URLs
- 05Bearer-token authentication
Use cases
- 01Search the registry for MCP servers by owner or repository
- 02Inspect a server's connection configuration
- 03Create a WebSocket URL for a configured server
MCP support
Exposes Smithery Registry API operations through an OpenAPI server, including MCP server discovery and WebSocket URL generation. It helps clients obtain launch configurations for MCP servers.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMESmithery Registry OpenAPI Server
This server provides an OpenAPI interface to the Smithery Registry API, allowing you to search and obtain launch configurations for Model Context Protocol (MCP) servers.
Features
- List available MCP servers with advanced filtering
- Get detailed server information including connection configurations
- Generate WebSocket URLs with base64-encoded configurations
- Authentication via bearer token
- Compatible with OpenAPI tools and clients
Authentication
All requests to the Smithery Registry API require authentication:
- You need a bearer token in the
Authorizationheader - Format:
Authorization: Bearer your-api-token - Obtain a token from your Smithery user profile under API keys
Usage
Installation
pip install -r requirements.txt
Running the server
uvicorn main:app --host 0.0.0.0 --port 8000 --reload
API Endpoints
-
GET /servers- List available MCP servers- Query parameters:
q: Search query (semantic search)page: Page number (default: 1)pageSize: Items per page (default: 10)
- Advanced filtering syntax in the
qparameter:- Text search: "machine learning"
- Owner filter: owner:username
- Repository filter: repo:repository-name
- Deployment status: is:deployed
- Example: "owner:smithery-ai repo:fetch is:deployed machine learning"
- Query parameters:
-
GET /servers/{qualified_name}- Get detailed information about a specific server -
POST /create-websocket-url- Create a WebSocket URL with encoded config- Request body:
{ "qualifiedName": "username/repository", "config": { // Configuration object matching the server's schema } }
- Request body:
WebSocket Connection
For direct WebSocket connections to MCP servers:
- URL Format:
https://server.smithery.ai/${qualifiedName}/ws?config=${base64encode(config)} - Config must comply with the server's configSchema
- Config is base64-encoded JSON
Integration with OpenAPI Tools
This server is designed to be compatible with any OpenAPI-compliant tools and frameworks. The API follows RESTful conventions and provides detailed OpenAPI documentation at the /docs endpoint when running.
TypeScript SDK Usage
For use with the Smithery TypeScript SDK:
import { WebSocketClientTransport } from "@modelcontextprotocol/sdk/client/websocket.js"
import { createSmitheryUrl } from "@smithery/sdk/config.js"
const url = createSmitheryUrl(
"https://your-smithery-mcp-server/ws",
{
// config object matching schema
},
)
const transport = new WebSocketClientTransport(url)
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.