About Peak Answer
Peak Answer is a remote service for measuring how AI search engines represent a brand and identifying site issues that can prevent answers from citing it. It is for brands tracking AI visibility, competitor wins and recommended content or technical actions.
Key features
- 01Audit site readiness for AI answers
- 02Measure brand visibility over time
- 03Track questions competitors win
- 04Generate content and semantic assets
- 05Review ranked actions and Search Console queries
Use cases
- 01Check whether a page can be quoted by AI search
- 02Find buying questions where competitors lead
- 03Track brand visibility across measured history
MCP support
The remote MCP server exposes read-only site audits, content generation and brand-specific visibility data to MCP clients. It uses OAuth 2.0 discovery or an API key as a bearer token.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://peakanswer.com/api/mcpclaude mcp add --transport http peakanswer-peak-answer https://peakanswer.com/api/mcp{"mcpServers":{"peakanswer-peak-answer":{"type":"http","url":"https://peakanswer.com/api/mcp"}}}Open in Cursor{"servers":{"peakanswer-peak-answer":{"type":"http","url":"https://peakanswer.com/api/mcp"}}}Open in VS Code{"mcpServers":{"peakanswer-peak-answer":{"type":"http","url":"https://peakanswer.com/api/mcp"}}}{"mcpServers":{"peakanswer-peak-answer":{"serverUrl":"https://peakanswer.com/api/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEPeak Answer MCP
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
claude mcp add --transport http peak-answer https://peakanswer.com/api/mcp
mcp.json (Cursor, VS Code and other clients that read this format)
{
"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 |
Every tool is read-only and annotated as such, so a client does not have to ask before each call.
audit_site, generate and geo_audit work on any public URL. The my_* tools read the
customer's own measured data.
Who it is for
Anyone with a Peak Answer account; the MCP is included on every plan. The free tools at peakanswer.com/tools need no account at all.
Agent skills that use these tools, and work without them: peakanswer/peak-answer-skills.
Registry
server.json in this repo is the MCP Registry entry.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.