Your firewall is your AI policy — I probed 18 major sites to read it

Everyone's arguing about AI search visibility. Almost nobody is measuring the boring part: whether AI crawlers can fetch your pages at all.

So I built a small open-source tool — — same engine, plus it shows you the exact text an AI crawler can extract from your page. And yes: the scanner site itself scores 100/100 on its own audit, because on the first run it scored 70 and I had to eat my own cooking.

How this was built (transparently)

I built this in one working session with Claude as the pair — spec, research, code, and validation loop. The part worth sharing: the first live run produced phantom "WAF block" findings that turned out to be transient connection failures. We caught it because the validation run was against 14 production sites I operate, where I could sanity-check every claim — and the fix (retry logic + a separate probe-error class that never masquerades as a finding) became the tool's most important feature. Before publishing this post, an adversarial fact-check against the dataset killed four more of my favorite sentences — including "Airbnb blocks AI crawlers," which turned out to be two-thirds true: it 403s the OpenAI, Perplexity, Microsoft, Amazon, Meta, and Common Crawl UAs while serving all three Anthropic ones a 200. The meta-skill isn't writing Python. It's knowing what the tool — and the write-up — must refuse to claim.

Tool: . Corrections welcome — especially from anyone at the companies named, with logs.