MCP servers overview

Aigeon exposes its capabilities through the Model Context Protocol (MCP) — an open standard that lets AI assistants like Claude call external tools directly. Instead of copy-pasting API calls, you can ask your AI agent to "send the weekly newsletter", "pull click rates for last month's campaign", or "add a contact to the beta list" — and it will call Aigeon on your behalf.

What is MCP?

MCP is a lightweight protocol that turns any service into a set of tools an LLM can discover and invoke. Your AI client connects to an MCP server (a small HTTP/stdio process), receives a manifest of available tools, and calls them when answering your requests.

Aigeon's MCP server wraps the same public REST API described in the API reference — the same permissions, the same rate limits, the same JSON envelope.

How to connect

You need:

1. An Aigeon API key (create one at Settings → API Keys).
2. An AI client that supports MCP — Claude Desktop, Claude Code CLI, or Cursor.
3. A one-time config edit to point your client at the Aigeon MCP server.

Step-by-step guides:

• Claude Desktop
• Claude Code
• Cursor

Security notes

• Your API key is stored locally in the client config file — never committed to version control.
• The MCP server only has the permissions your API key grants. Use the narrowest scope that fits your workflow (e.g. sending_access for a "send newsletter" agent, analytics_access for a reporting agent).
• All tool calls go through Aigeon's standard auth middleware — the AI client cannot do anything you couldn't do with the REST API directly.