Lagotto Meter checks if your llms.txt, llms-full.txt and JSON-LD actually support what your site claims. Not an SEO tool — it reads only what you've declared for AI agents and scores the gap between your claims and your proof. Runs on Llama 3.3 70B with Gemini 2.5 Flash as fallback. Free, 1 scan/IP/24h, multilingual (EN/IT/DE/ES). The scoring prompt is public — nothing hidden, including the rule that penalized our own site.

Hey Product Hunt 👋 Maker here. Built this after getting frustrated watching AI agents confidently hallucinate things about companies that simply weren't true — not because the AI was "wrong", but because the site never declared the facts anywhere a crawler could read them. Lagotto Meter doesn't read your homepage copy. It only reads what you've structured for AI agents: llms.txt, llms-full.txt, JSON-LD. If your site claims something your structured data doesn't support, you get penalized — even if the claim is true. The point isn't to catch liars, it's to push everyone toward declaring facts explicitly instead of hoping an LLM infers them correctly. Heads up: it's rate limited to 1 scan per IP per day, so pick your URL wisely before hitting analyze 😄 Happy to answer anything about the architecture, the scoring logic, or why we chose to penalize ourselves too. 🐕

Love the transparency move with the public scoring prompt, especially calling out that it penalized their own site first. That's the kind of honest craft that makes me actually trust the score I'm getting back.

@ceylani0pt Thank you! That was a deliberate choice — if the tool couldn't be honest about its own creator's site, the scores would mean nothing. BASIC (our Lagotto Romagnolo) approved the methodology. 🐾

Lagotto Meter

Check what AI agents actually understand about your site

Check what AI agents actually understand about your site