Claude: Your Writer, Analyst and Assistant

Claude is an AI model made by Anthropic. Think of it as a very fast, very well-read assistant that’s especially good at careful writing, reading long documents and working through tasks step by step.

What we use it for

  • Drafting customer replies, proposals and newsletters in your voice.
  • Reading contracts, reports and spreadsheets and summarizing what matters.
  • Running multi-step jobs in your command center, like “check today’s orders, flag problems and draft the emails”.
  • Writing video scripts that ElevenLabs, HeyGen or Synthesia turn into finished videos.
Screenshot: Claude drafting a customer reply using order details from the store
Claude drafting a reply with the real order details, ready for approval.

Why it fits our approach

Anthropic created the Model Context Protocol, the open standard we use to connect AI to your systems. That means Claude can safely use your website, store and books through the same connections any other AI could use, so you’re never locked in to one company.

GIF: Claude connected to WordPress, listing recent orders and drafting a summary
Connected through open standards, not a one-off integration.

Good to know

  • Business and API plans let you control how your data is handled. We set this up on accounts in your name.
  • It’s not always the cheapest option for simple, high-volume jobs. For those we may use a smaller model.
Go deeper (a little techy)

We use Claude through the API and through MCP-capable clients. Your systems are exposed as MCP servers (WordPress MCP Adapter, Directus MCP, ERPClaw), and Claude calls their tools with scoped credentials. Prompts, tool definitions and logs are stored in your repository and database so the model can be swapped.

Google these: “Claude API”, “Anthropic Model Context Protocol”, “Claude tool use”, “Claude Code”.