Word of Agent

ConnectorRemote MCP server, OAuth

Word of Agent in your assistant

Word of Agent measures whether search engines and AI assistants can find, read and recommend a business, and says what to change. Connected to your assistant, it gives the assistant the measurements and the work: your assistant fixes, writes and drafts; you approve anything that goes out in your name; Word of Agent checks the result.

Connect it

The server address is https://app.wordofagent.app/mcp. You sign in to Word of Agent and choose which account the assistant may use.

  • Claude: Settings, Connectors, Add custom connector, paste the address, then Connect.
  • ChatGPT: Settings, Apps and Connectors, Create (turn on developer mode under Advanced if you do not see it), and use the same address.
  • Claude Code, Codex and other agents: add it as a remote MCP server with that address; they open the sign-in in your browser.

A free account is enough to start: sign up with your email and add your site.

What your assistant can do with it

  • Measure. Audit a site: crawler access, indexing in Google and Bing, rendering, structured data, the facts the site states, listings and reviews. Read the five outcome lines: found in search, recommended by assistants, talked about, trusted, heard of.
  • Ask the customers' questions. Get the questions, answer them the way it would answer anyone, and send the answers; read them back for who is named instead, what is said wrong, and why.
  • Fix. Work through the to do list most costly first, with the exact fix for each finding, then have Word of Agent check it again.
  • Get the business heard. A main channel chosen from evidence and your limits, pages to write, sources to get mentioned on, local listings, reviews, rivals compared, listening, and for businesses that sell to businesses, careful cold email.
  • Draft for you to approve. Outreach, replies, review requests and pitches are drafts: nothing is sent, posted or published until you approve it in Word of Agent, and then you or your assistant sends it.

What it never does

  • It calls no AI model: your own assistant does the thinking, on the plan you already pay for.
  • It sends nothing in your name, and refuses drafts that break the rules for their kind: paying for mentions, incentives for reviews, undisclosed community posts, cold email without an opt-out or where the law needs consent.
  • It does not read Reddit or X, scrape personal data or buy lists.
  • An assistant cannot approve its own drafts or plans: only a person signed in to Word of Agent can.

Try asking

  • "Set up Word of Agent for example.com: draft my customers' questions with me, then collect the first AI answers and tell me what you found."
  • "Why don't AI assistants recommend example.com, and what should we fix first?"
  • "Work through the Word of Agent to do list for example.com, most costly first."
  • "What's my main channel this month, and what are the first three things to do?"

Data and support

What we collect and why is in the privacy policy; the terms cover using the service. Questions or problems: [email protected].