Peak Answer MCP

Name: Peak Answer
Author: peakanswer

The remote MCP server for Peak Answer: whether AI search engines recommend a brand, which buying questions competitors win instead, and the technical and content reasons an engine cannot quote a site.

Remote only. There is nothing to install and no local process; you point a client at the URL.

https://peakanswer.com/api/mcp

Connect

Claude Code

bash

claude mcp add --transport http peak-answer https://peakanswer.com/api/mcp

mcp.json (Cursor, VS Code and other clients that read this format)

json

{
  "mcpServers": {
    "peak-answer": {
      "url": "https://peakanswer.com/api/mcp"
    }
  }
}

Claude.ai and ChatGPT — add the URL under Settings, Connectors.

Setup with screenshots: https://peakanswer.com/tools/claude-mcp

Authentication

OAuth 2.0, discovered automatically. An unauthenticated request returns 401 with

WWW-Authenticate: Bearer resource_metadata="https://peakanswer.com/.well-known/oauth-protected-resource/api/mcp", scope="mcp"

and the client takes it from there. Clients that would rather hold a token can use an API key from Peak Answer Settings as a bearer token instead.

The connection is scoped to one brand. Call my_brand to find out which domain that is rather than asking the user or guessing.

Tools

Tool	What it does
`audit_site`	Answer readiness, AI crawler access, robots.txt, sitemap, structured data, headings, meta description, readability and entities for one URL, in one call. Seconds
`generate`	Produces `llms_txt`, `faq_schema`, `content_brief`, `prompt_set` or `semantic_keywords`
`geo_audit`	Asks a real model ten real buying questions. Starts a background run and returns immediately; call again with the same domain to collect it
`my_brand`	Which domain this connection is scoped to
`my_visibility`	Measured visibility history for that brand, rather than a one-off check
`my_questions`	Tracked questions, with `only_losing` for the ones a competitor wins
`my_question_history`	One question over time
`my_actions`	The ranked backlog for that brand
`my_search_console`	Search Console queries near the top of page two