@agentpowers/mcp-server

MCP server for the AgentPowers marketplace. Search, browse, install, and manage paid skills and agents directly inside Claude Code and Claude Cowork conversations.

Tools

| Tool | Description | |------|-------------| | search_marketplace | Search for skills and agents by query, category, or type | | get_skill_details | Get detailed info (price, security status, platforms, versions) for a skill | | install_skill | Download and install a skill after security check | | check_purchase_status | Check payment status and retrieve license code | | check_installed | List all installed skills with version and edit status | | uninstall_skill | Remove an installed skill and its version pin | | check_for_updates | Check installed skills for newer versions | | get_categories | List all marketplace skill categories | | get_skill_reviews | Fetch reviews for a specific skill | | start_checkout | Begin the purchase flow for a paid skill (returns checkout URL) | | get_account_profile | Retrieve the authenticated user's account profile | | list_purchases | List all skills purchased by the authenticated user |

Setup

Add to your Claude Code MCP configuration (~/.claude/claude_desktop_config.json):

{
  "mcpServers": {
    "agentpowers": {
      "command": "npx",
      "args": ["-y", "@agentpowers/mcp-server"]
    }
  }
}

Or install globally first:

npm install -g @agentpowers/mcp-server

Then configure with the global path:

{
  "mcpServers": {
    "agentpowers": {
      "command": "agentpowers-mcp"
    }
  }
}

Authentication

Most tools work without authentication. Installing paid skills and checking purchase status require a CLI auth token stored at ~/.agentpowers/auth.json. Authenticate via:

npx @agentpowers/cli login

Or if you have the CLI installed:

ap login

Configuration

The server connects to https://api.agentpowers.ai by default. Override with the AGENTPOWERS_API_URL environment variable for development.

Security

Every skill listed on the marketplace passes a multi-layer security pipeline before listing, including static validation, dependency scanning, malware detection, and AI-powered security review. Security results are displayed in skill details.

License

MIT