Get a weekly AI visibility report in Claude

SearchSeal has a saved workflow called weekly_pulse. Claude runs it against your account: it compares the last seven days with the seven before, per AI tool, names the rival that gained most and lists your three highest-impact fixes in under 200 words. You connect once with a read-only token.

Check your site
Evidence trail
  • Seven days against the seven before
  • Visibility and share of voice per AI tool
  • The three highest-impact fixes

Run the weekly report in six steps

  1. Create a Read token. In the SearchSeal app open Settings, then API, and create a personal access token with Read scope. A read token cannot change questions or fixes.
  2. Connect Claude Code. Add the server with your token in place of <token>. Keep the finished command out of shared shell history.
    claude mcp add searchseal --transport http https://searchseal.com/api/v1/mcp --header "Authorization: Bearer <token>"
  3. Or connect Claude Desktop. Claude Desktop needs the local bridge configuration from the MCP setup guide, because SearchSeal accepts a Bearer token and has no OAuth sign-in. Reload the app after saving the configuration.
  4. Find your brand id. Ask Claude to list your brands. Each saved workflow needs a brand id.
  5. Run weekly_pulse. Choose weekly_pulse from the saved workflows and enter the brand id. If your client does not list MCP prompts, ask in plain words for the last seven days against the seven before, by AI tool.
  6. Read the result, then ask follow-ups. Open the questions behind any change by asking Claude for the saved answers. Repeat the run next week.

What the workflow does

weekly_pulse gives Claude a fixed instruction. Claude calls the brand report for the last seven days with the AI tool as the dimension, then the same report for the seven days before. It summarizes visibility and share of voice per AI tool, what rose or fell and which rival gained most. It then reads your fix list and names the three highest-impact fixes. The instruction caps the answer at 200 words.

The numbers come from saved answers to your tracked questions. Visibility is the share of those answers that name your brand.

How it differs from the Monday email

SearchSeal already sends a weekly report by email on Mondays, to the account owner and every team member. The Claude version runs when you ask. It suits a moment when you want to question the data: ask which rival gained most, then ask for that rival's saved answers.

Claude does not run on a schedule by itself. You start the report each week, or use whatever scheduling your client offers.

Limits and privacy

Data you request goes to the AI client you connect, and that provider handles it under its own terms. Revoke the token in Settings, then API, to stop access at once.

Saved answers and source text are flagged as untrusted content, so treat them as data to read. Answers vary between runs, so a weekly change is a difference between two samples. A change after a page edit does not prove the edit caused it. Search data has its own data-through date and arrives a few days late.

See Connect Claude Code to SearchSeal for the full setup and the MCP setup guide for the other clients.

FAQ

Does weekly_pulse need a write token?

No. It only reads. A read token is enough, and write access is needed only to add questions or change fixes.

Can Claude send the report by email?

Not through SearchSeal. SearchSeal sends its own weekly email on Mondays. Claude returns the weekly_pulse summary in the chat.

Which plans include MCP access?

Every plan includes the REST API and the MCP server, with the same tokens and request limits.

Does the report cover every AI tool I track?

It reports per AI tool for the engines your plan runs. The report covers saved answers, so a tool with no answers in the period has nothing to compare.

Find your next AI search win

Check your site