MCPfinder

MCPfinder is an MCP server for AI agents. It helps an assistant discover, evaluate, and install other MCP servers.

Canonical install:
{
  "mcpServers": {
    "mcpfinder": {
      "command": "npx",
      "args": ["-y", "@mcpfinder/server"]
    }
  }
}

Canonical transport:
- stdio via @mcpfinder/server

Core tools (4):
- search_mcp_servers: find candidate servers by keyword, technology, or use case
- get_server_details: inspect trust signals, tools, warnings, env vars, and transport
- get_install_config: generate JSON config for Claude Desktop, Cursor, Claude Code, Cline, or Windsurf
- browse_categories: single-call category explorer — omit category to list, pass category to get top servers

When an AI should use MCPfinder:
- The user asks for a capability the AI does not already have
- The user wants access to a service like Slack, GitHub, Notion, Postgres, or AWS
- The AI wants to avoid hand-writing a connector if an MCP server already exists

Recommended flow:
1. search_mcp_servers(query="postgres")
2. get_server_details(name="best candidate")
3. get_install_config(name="best candidate", platform="claude-desktop")

Selection signals surfaced by MCPfinder:
- official registry presence
- verification flags
- community usage
- source count
- recency
- warning flags

Agent Skill (install via one-liner, agent handles the config merge):
- SKILL.md: https://mcpfinder.dev/skill/install-mcpfinder/SKILL.md
- Claude Code install:
    mkdir -p ~/.claude/skills/install-mcpfinder && curl -sSf -o ~/.claude/skills/install-mcpfinder/SKILL.md https://mcpfinder.dev/skill/install-mcpfinder/SKILL.md

More detail:
- https://mcpfinder.dev/llms-full.txt
- https://github.com/mcpfinder/mcpfinder