Skip to main content

Build with AI

Every action in Reactor is also a Model Context Protocol (MCP) tool. That means an AI agent — Claude Code, Claude Desktop, Cursor, or anything that speaks MCP — can drive Reactor end to end with no UI:

  • Configure the engine — set brand voice, positioning, writing rules, audiences, competitors, and per-platform tone.
  • Generate content — create briefs and produce platform-native drafts for X, LinkedIn, blog, newsletter, Discord, and more.
  • Publish and schedule — post now or queue to the calendar; reschedule and cancel.
  • Analyze — read cross-platform analytics, top content, insights, and posting recommendations, then feed them back into the next brief.

Two ways to start​

1. Hand setup to an agent​

Give an AI agent the Reactor setup skill and it will install the MCP, help you connect your services, and draft your initial engine content from an interview or your website.

→ Hand setup to an AI

2. Call the MCP directly​

Install the Reactor MCP in your client and start issuing tool calls yourself.

  • Install the MCP — add Reactor to Claude Code, Claude Desktop, or Cursor.
  • Authentication — generate an API key and pass it as a Bearer token.
  • MCP overview — how the server is structured and what the tools cover.

The server is self-describing​

There is no per-tool reference to memorize. Reactor's MCP server advertises its tools over the protocol — your client reads each tool's name, description, and input schema at connect time. See MCP overview for the tool categories and example workflows for end-to-end recipes.