What problem does it solve?
Finding, connecting, and using AI tools and skills scattered across many providers becomes slow, repetitive, and hard to manage.
Core Features & Use Cases
- MCP server discovery and connections: Search for MCP servers and create managed, long-lived connections that handle auth and session lifecycle.
- Tool discovery and execution: List tools, inspect schemas for inputs/outputs, and call tools by connection and name.
- Skill browsing and installation: Browse the Smithery Skills Registry and install skills with the upstream installer to enable canonical workflows like integrations and common agent actions.
- Example use case: You want to create GitHub issues from your agent—first connect to an MCP server, then inspect the tool schema, then call the tool with the required JSON payload.
Quick Start
Ask an AI assistant to help you run smithery mcp search "github" to find available MCP servers for GitHub integrations.