@aisearchready/mcp

The first AEO-audit Model Context Protocol server. Audit any URL for AI-search readiness from inside Claude Desktop, Cursor, LM Studio, or any MCP-compatible client.

Powered by aisearchready.io — the open standard for measuring how AI search engines (ChatGPT, Perplexity, Claude, Gemini, AI Overviews) read, crawl, and cite your site.

What it does

Three tools, no API key required for the free preview:

| Tool | What it returns | |---|---| | audit_url(url, deep?) | 135-signal AEO audit · score · critical issues · paste-ready fixes | | verify_cert(domain) | Canonical signed cert state for a certified domain | | get_standard() | The Open AEO Standard v1.0 methodology summary |

You can say things like:

"Audit example.com for AI search readiness"

"Is anthropic.com AI Search Ready certified?"

"What does the AEO score actually measure?"

…and the agent will call the right tool inline. No tab-switching, no sign-up, no rate-card surprises.

Install — Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "aisearchready": {
      "command": "npx",
      "args": ["-y", "@aisearchready/mcp"]
    }
  }
}

Restart Claude Desktop. You should see aisearchready in the MCP tools list (the hammer icon at the bottom of the input box).

Install — Cursor

1. Open Cursor → Settings → MCP
2. Click Add Server
3. Command: npx -y @aisearchready/mcp
4. Save and reload

Install — LM Studio, Continue.dev, Zed

The MCP transport is stdio, no special flags. Point your client at npx -y @aisearchready/mcp as the command.

Studio API key (optional)

The free tier returns scan IDs and a link to the full preview report. A Studio API key (sk_studio_…) unlocks:

• Full structured score + pillar breakdown
• Every issue with priority + paste-ready fixCode
• Deep Audit (multi-page site sample)
• Citation Probe outcomes (Perplexity + Claude + OpenAI)

Set AISEARCHREADY_API_KEY in your client's MCP env block:

{
  "mcpServers": {
    "aisearchready": {
      "command": "npx",
      "args": ["-y", "@aisearchready/mcp"],
      "env": {
        "AISEARCHREADY_API_KEY": "sk_studio_…"
      }
    }
  }
}

Get a key at aisearchready.io/pricing (£199/mo Studio · 250 audits + Citation Probe + white-label included).

Develop locally

git clone https://github.com/ed2903-web/aisearchready.git
cd aisearchready/mcp-server
npm install
npm run dev   # tsx ./src/index.ts — runs the MCP server on stdio

Point it at a local app instead of production:

AISEARCHREADY_BASE_URL=http://localhost:3000 npm run dev

Source + licence

• Repository: github.com/ed2903-web/aisearchready (the whole audit engine; this package lives in mcp-server/)
• Methodology: aisearchready.io/standard
• Licence: MIT

Why this exists

The category of AEO audit tools is consolidating fast. Profound, Athena, Otterly, Semrush AI Visibility, Brand Radar — every one of them is a dashboard you log into. None of them work inside the agent ecosystem where the buyer increasingly lives.

An MCP server reverses the polarity: the audit comes to you, in the AI chat where you're already working. "Audit example.com" becomes a one-line natural-language command. The brand impression is in the chat, not on a marketing site you have to remember to visit.

Per Screaming Frog's MCP launch (May 2026), the category is moving. This is the first MCP server purpose-built for AEO measurement.

See the Open AEO Standard v1.0 for the methodology and the reviewer-signature call.