MCP server

The Model Context Protocol lets AI assistants call external tools. With the Pidgely MCP server, an assistant can send emails and work with your Pidgely resources (emails, domains, contacts, broadcasts) using your API key. It runs locally over stdio with npx and reads the key from PIDGELY_API_KEY.

#Claude Desktop

Add the server to claude_desktop_config.json (Settings → Developer → Edit config) and restart Claude Desktop:

{
  "mcpServers": {
    "pidgely": {
      "command": "npx",
      "args": ["-y", "@pidgely/mcp"],
      "env": { "PIDGELY_API_KEY": "pdg_your_key_here" }
    }
  }
}

#Claude Code

claude mcp add --transport stdio pidgely --env PIDGELY_API_KEY=pdg_your_key_here -- npx -y @pidgely/mcp

#Cursor

Add it to .cursor/mcp.json in your project (or ~/.cursor/mcp.json for every project):

{
  "mcpServers": {
    "pidgely": {
      "command": "npx",
      "args": ["-y", "@pidgely/mcp"],
      "env": { "PIDGELY_API_KEY": "pdg_your_key_here" }
    }
  }
}

#Choosing a key for an agent

An agent acts with the permissions of its key. Start narrow:

• For agents that only send, use a sending_access key restricted to one domain: it can't read contacts, change domains or create keys.
• Use a full_access key only for agents you trust to manage the account, and revoke it when you're done.
• Ask the assistant to confirm before sending to real recipients.

#Docs for assistants

Assistants can also read these docs directly: /llms.txt lists every page, /llms-full.txt has all of them in one file, and every page is available as Markdown at <url>.md.