Agent-Readiness Audit

Field audit for agent access

Can an AI agent actually find your pricing, sign up, or contact you? We test it, fix it, and prove it.

Paste a domain. A free check tries pricing, sign-up, and contact three times and keeps a screenshot of where a failure stopped. A paid check runs five times. The 0–100 score summarizes that proof.

One audit fetches about twenty URLs and opens the homepage in a headless browser. Each request stops at 10 seconds.

Or browse the Agent-Readiness Index of well-known software companies. Agencies can ask for done-for-you.

The screenshots

The free teaser shows whether pricing, sign-up, and contact were found. The full report shows the pages the agent opened, with the screenshot for each task.

The full report

The report lists every check, the pass rate for each task, screenshots of where a failure stopped, and who ran the test. It opens after a $49 Stripe payment. The webhook is what unlocks it. A $29 monthly monitor re-runs the three tasks on a schedule and emails you when a pass rate drops.

The Fix Pack

$199 once. After payment you get llms.txt, robots.txt rules for AI agents, a starter OpenAPI spec, an MCP manifest, and paste-in links for any task the agent missed. Verify fixes runs each task five times and shows the pass rate before and after.

What the score is made of

Pass earns the full weight, a warning earns half, and a failure earns nothing. Discover, access, and tasks are 80 points. An API or a payment challenge is the other 20.

  1. Tasks50 ptsPass rate for finding pricing, sign-up, and contact. Half the score.
  2. Access20 ptsAI crawlers are allowed, agent user agents are not blocked, and the homepage is not a challenge.
  3. Content10 ptsThe raw HTML, before JavaScript runs, contains text an agent can read.
  4. API12 ptsA remote MCP server that answers initialize and tools/list, or an OpenAPI file.
  5. Bonus8 ptsllms.txt and an HTTP 402 payment challenge. Small on purpose.