Smithery logo

Smithery

Smithery is a platform for discovering, deploying, and managing Model Context Protocol (MCP) servers and skills. It operates a public registry of community-built MCP extensions that AI agents can use to access external tools, data sources, and services, plus a Connect gateway that bundles connections in a namespace behind a single MCP endpoint at mcp.smithery.run. The platform exposes APIs for server registry browsing, server deployment, skill publishing, namespace management, scoped service tokens, connection lifecycle, trigger subscriptions, and an MCP transport endpoint for AI-agent integration.

1 APIs 0 Features
Artificial IntelligenceLarge Language ModelsMCPModel Context ProtocolAI AgentsDeveloper ToolsRegistrySkillsTool Discovery

APIs

Smithery Platform API

The Smithery Platform API provides programmatic access to the MCP server registry and the Connect gateway. It groups 55 operations across 35 paths under eight tags (servers, ski...

Semantic Vocabularies

Smithery Context

25 classes · 0 properties

JSON-LD

API Governance Rules

Smithery API Rules

8 rules · 2 errors 6 warnings

SPECTRAL

Resources

🔗
LinkedIn
LinkedIn
👥
GitHubOrganization
GitHubOrganization
🔗
Documentation
Documentation
🔗
CLI
CLI
🔗
Website
Website
🔗
Playground
Playground
📰
Blog
Blog
🔗
Documentation
Documentation
🔑
Authentication
Authentication
📦
SDK
SDK
📦
SDK
SDK
💻
SampleCode
SampleCode
🔗
Documentation
Documentation
🔗
Documentation
Documentation
🔗
Documentation
Documentation
🔗
Documentation
Documentation
🔗
Documentation
Documentation
🔗
Documentation
Documentation
🔗
Documentation
Documentation
💻
SampleCode
SampleCode
🔗
Integration
Integration
🔗
MCPServer
MCPServer
🔧
Tools
Tools
🔗
MCPServer
MCPServer
🔧
Tools
Tools
🔧
Tools
Tools
🔧
Tools
Tools

Sources

Raw ↑
aid: smithery
name: Smithery
description: >-
  Smithery is a platform for discovering, deploying, and managing Model Context Protocol (MCP) servers and skills. It
  operates a public registry of community-built MCP extensions that AI agents can use to access external tools, data
  sources, and services, plus a Connect gateway that bundles connections in a namespace behind a single MCP endpoint at
  mcp.smithery.run. The platform exposes APIs for server registry browsing, server deployment, skill publishing,
  namespace management, scoped service tokens, connection lifecycle, trigger subscriptions, and an MCP transport
  endpoint for AI-agent integration.
type: Index
position: Consuming
access: 3rd-Party
image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
tags:
  - Artificial Intelligence
  - Large Language Models
  - MCP
  - Model Context Protocol
  - AI Agents
  - Developer Tools
  - Registry
  - Skills
  - Tool Discovery
created: '2025-08-19'
modified: '2026-05-22'
url: https://raw.githubusercontent.com/api-evangelist/smithery/refs/heads/main/apis.yml
specificationVersion: '0.19'
apis:
  - aid: smithery:platform-api
    name: Smithery Platform API
    description: >-
      The Smithery Platform API provides programmatic access to the MCP server registry and the Connect gateway. It
      groups 55 operations across 35 paths under eight tags (servers, skills, tools, tokens, namespaces, organizations,
      connect, connect.mcp), covering registry search, server CRUD, releases and runtime logs, secrets, managed domains,
      icon management, skill bundles, organization API keys, namespace administration, scoped service tokens, and the
      Connect API for managing connections, triggers, and subscriptions plus a single POST
      /connect/{namespace}/{connectionId}/mcp endpoint that proxies MCP streamable-HTTP traffic to upstream servers.
    humanURL: https://smithery.ai/docs
    baseURL: https://api.smithery.ai
    tags:
      - Artificial Intelligence
      - MCP
      - Registry
      - Servers
      - Skills
      - Tools
      - Connect
      - Tokens
      - Namespaces
      - Organizations
    properties:
      - type: Documentation
        url: https://smithery.ai/docs
      - type: OpenAPI
        url: https://raw.githubusercontent.com/api-evangelist/smithery/refs/heads/main/openapi/smithery-openapi.json
      - type: OpenAPI
        url: >-
          https://raw.githubusercontent.com/api-evangelist/smithery/refs/heads/main/openapi/smithery-documented-openapi.yml
      - type: OpenAPI Specification
        url: https://smithery.ai/docs/openapi.json
      - type: OpenAPI Specification
        url: https://app.stainless.com/api/spec/documented/smithery/openapi.documented.yml
      - type: llms.txt
        url: https://smithery.ai/docs/llms.txt
      - type: APIReference
        url: https://smithery.ai/docs/api-reference
      - type: JSONSchema
        url: json-schema/smithery-server-schema.json
      - type: JSONSchema
        url: json-schema/smithery-skill-schema.json
      - type: JSONStructure
        url: json-structure/smithery-server-structure.json
      - type: JSONLD
        url: json-ld/smithery-context.jsonld
      - type: SpectralRules
        url: rules/smithery-rules.yml
      - type: Vocabulary
        url: vocabulary/smithery-vocabulary.yml
      - type: Plans
        url: plans/smithery-plans-pricing.yml
      - type: RateLimits
        url: rate-limits/smithery-rate-limits.yml
      - type: FinOps
        url: finops/smithery-finops.yml
common:
  - type: LinkedIn
    url: https://www.linkedin.com/company/smithery-ai
  - url: https://github.com/smithery-ai
    name: GitHub Organization
    type: GitHubOrganization
  - name: Smithery Documentation
    description: Complete API documentation and developer guides
    url: https://smithery.ai/docs
    type: Documentation
  - name: Smithery CLI
    description: >-
      Command-line interface for managing MCP servers and skills (npm install -g smithery@latest, Node.js 20+, AGPL-3.0,
      725 stars)
    url: https://smithery.ai/docs/concepts/cli
    type: CLI
  - name: Smithery Website
    description: Main platform for discovering and using MCP extensions
    url: https://smithery.ai/
    type: Website
  - name: Smithery Playground
    description: Interactive playground for testing MCP servers
    url: https://smithery.ai/playground
    type: Playground
  - name: Smithery Blog
    description: Blog covering MCP, AI agents, and platform updates
    url: https://smithery.ai/blog
    type: Blog
  - name: Smithery Skills Directory
    description: Browse reusable skills and prompt-based AI capabilities
    url: https://smithery.ai/skills
    type: Documentation
  - name: API Keys
    description: Manage API keys for Smithery platform access
    url: https://smithery.ai/account/api-keys
    type: Authentication
  - name: TypeScript SDK
    description: >-
      @smithery/api - Server-side TypeScript / JavaScript client (generated by Stainless), full type definitions,
      auto-retry, file upload, auto-pagination, multi-runtime (Node, Deno, Bun, Cloudflare Workers, Vercel Edge)
    url: https://github.com/smithery-ai/typescript-api
    type: SDK
  - name: Smithery CLI GitHub Repository
    description: >-
      Smithery CLI source — Install, manage and develop MCP servers and skills for agents (TypeScript, AGPL-3.0, 725
      stars)
    url: https://github.com/smithery-ai/cli
    type: SDK
  - name: Smithery Cookbook
    description: Collection of recipes showcasing effective ways of using Smithery
    url: https://github.com/smithery-ai/smithery-cookbook
    type: SampleCode
  - name: MCP Concepts Documentation
    description: 'Conceptual primer: What is MCP, Namespaces, CLI usage'
    url: https://smithery.ai/docs/concepts/what_is_mcp
    type: Documentation
  - name: Connect API — Deep Linking
    description: Deep-link patterns for routing users into a Connect flow
    url: https://smithery.ai/docs/use/deep-linking
    type: Documentation
  - name: Connect API — Token Scoping
    description: Documented token scopes for Connect-issued service tokens
    url: https://smithery.ai/docs/use/token-scoping
    type: Documentation
  - name: Connect API — Uplink
    description: Uplink lets locally-running MCP servers be exposed through Smithery without a public deployment
    url: https://smithery.ai/docs/use/uplink
    type: Documentation
  - name: Connect API — Listing Your Client
    description: How AI-agent clients list themselves on Smithery
    url: https://smithery.ai/docs/use/listing_your_client
    type: Documentation
  - name: Build — Publish MCP Server
    description: 'Publishing workflow: URL-based proxying or local MCPB bundles; auto-scan via /.well-known/mcp/server-card.json'
    url: https://smithery.ai/docs/build/publish
    type: Documentation
  - name: Build — Triggers
    description: Build server-side triggers that fire connection-scoped events
    url: https://smithery.ai/docs/build/triggers
    type: Documentation
  - name: Cookbook — TypeScript OAuth Client
    description: Example for building an OAuth client against Smithery in TypeScript
    url: https://smithery.ai/docs/cookbooks/typescript_oauth_client
    type: SampleCode
  - name: Integration — Vercel AI SDK
    description: Integration recipe for using Smithery with the Vercel AI SDK
    url: https://smithery.ai/docs/integrations/vercel_ai_sdk
    type: Integration
  - name: Smithery MCP Server (mcp.smithery.run)
    description: Hosted MCP gateway endpoint that bundles all connections in a namespace behind a single URL
    url: https://mcp.smithery.run
    type: MCPServer
  - name: agent.pw
    description: 'Smithery-built tool: Share APIs with agents, without sharing secrets'
    url: https://github.com/smithery-ai/agent.pw
    type: Tools
  - name: mouseless
    description: 'Smithery-built tool: Rust MCP server for macOS desktop control'
    url: https://github.com/smithery-ai/mouseless
    type: MCPServer
  - name: mcp-to-cli
    description: 'Smithery-built tool: Convert MCP servers to CLI tools'
    url: https://github.com/smithery-ai/mcp-to-cli
    type: Tools
  - name: workers-biscuit
    description: 'Smithery-built tool: Biscuit tokens for Cloudflare Workers'
    url: https://github.com/smithery-ai/workers-biscuit
    type: Tools
  - name: registry
    description: 'Smithery-built repo: "The fastest way to extend your AI"'
    url: https://github.com/smithery-ai/registry
    type: Tools
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com