The free scan carries the whole idea: three scores, the gaps in how AI currently describes the brand, and the issues worth fixing. The paid tools then break the score down, benchmark against a competitor, and hand back code.
A hosted MCP server that audits a website for AI search visibility. A scan returns three scores out of 100 — AEO, GEO and Agent Readiness — plus a mention readiness score, an AI Identity Card showing how AI currently describes the brand, and an auto-detected business type. Three deeper tools audit, compare and fix.
- A triple-score scan of any URL, with top issues and the business type it detected
- An AI Identity Card: how AI describes the brand today, its gaps, its citable claims and the competitors it names
- A full breakdown of 58+ checks across 12 categories
- A two-site gap analysis with category winners and overtake fix code
- Generated fix code with two-tier score projections, applied directly from a coding client
- A prompt that walks the scan, compare, audit and fix sequence in order
scan_site needs nothing at all — no API key, no wallet — though it is rate limited per IP and per URL. audit_site, compare_sites and fix_site need an API key from scan.convrgent.ai, or pay-per-call in USDC over x402 on Base, each priced separately. Connect remotely at https://aeo-mcp.convrgent.ai/mcp, or run it locally with uvx aeo-scanner.
One command plus a key — claude mcp add --transport http aeo-scanner https://aeo-mcp.convrgent.ai/mcp, then supply credentials