ChatGPT, Claude, Perplexity and Gemini increasingly visit your site through machine-readable files before they ever parse the HTML. On October 6, 2026 we scanned 20 Calgary-area plumbing & HVAC company websites. None graded C or better; the best scored 49 out of 100.
Scanned 2026-10-06 with agentready, one pass. 6 of the 20 sites could not be read by the
scanner (3 bot walls, 2 domains with no site connected, 1 redirect loop); they are still counted, and the 5 that
answered are scored on what an agent actually received. Full leaderboard.
| Site | Score | What it's missing |
|---|---|---|
| Site A | 49 | AI crawler rules in robots.txt, agent manifest, markdown |
| Site D | 49 | AI crawler rules in robots.txt, agent manifest, markdown |
| Site F | 37 | no llms.txt, no AI crawler rules |
| Site N | 10 | no llms.txt, no structured data, no meta description |
| Site R | 4 | not readable: no site connected to the domain |
Your score, what AI assistants currently can't see, and the exact files to publish. Three pages, delivered in 48 hours.
We write the llms.txt, add schema.org markup, configure AI crawler rules, and re-scan to prove the score moved.
We re-scan your site every week and tell you when your score drops or a competitor passes you.
agentready — 15 checks, 93 weighted points, zero dependencies.
Every number on this page is reproducible by re-running the scan (figures from 2026-10-06; live sites change, so expect drift).