What's new

Improvements and fixes, 29 October to 2 November

Your assistant can read Sources for one engine, list the prompts citing each page and see excerpts around each name. The MCP has fair-use limits, and the product is now called Firemark everywhere.

ImprovedMCPSourcesPromptsFan-out

  • get_sources takes an engine, so your assistant can read one engine's sites, types and your pages on that engine's answers.
  • Each cited page lists the prompts whose answers cite it, by engine, and list_prompts can show only the prompts that cite one page.
  • list_prompts says which way a changed verdict went: from named to not named, or back.
  • get_fanouts lists the ChatGPT answers where one of your pages came back and the answer still did not name you.
  • get_prompt and get_gaps include short excerpts around your brand and each competitor an answer names.
  • The MCP server has fair-use limits on calls a minute. Its version is now 1.2.0.
  • Firemark AI is now called Firemark in the app, in emails and in your assistant's connection.

If your assistant does not show the new tools yet, start a new chat or reconnect Firemark in its settings.

Find out if AI recommends you

Start tracking your brand, or open the demo to see a full report first.