Agent Web Index › data.gouv.fr
Can AI assistants read data.gouv.fr?
Measured on 2026-09-19 by asking the site 7 times — once as a browser, once as each of the 6 crawlers that feed ChatGPT, Claude, Perplexity, Gemini, Meta AI, Apple Intelligence and Doubao — and comparing what came back. The AI-training opt-out tokens that never crawl are read from robots.txt instead.
B75 / 100
6 of 6 crawlers can read this page.
More readable than 52% of the 47,298 sites measured so far.
No JSON-LD.
What each crawler got back
| Crawler | HTTP | Result | robots.txt |
|---|---|---|---|
| ClaudeBot (Claude) | 200 | can read it | allowed-by-star |
| GPTBot (ChatGPT) | 200 | can read it | allowed |
| OAI-SearchBot (ChatGPT Search) | 200 | can read it | allowed-by-star |
| PerplexityBot (Perplexity) | 200 | can read it | allowed-by-star |
| Google-Extended (Gemini, AI Overviews) a robots.txt token, not a crawler: it controls how already-crawled pages may be used, and never makes a request of its own | — | not opted out | allowed-by-star |
| Meta-ExternalAgent (Meta AI) | 200 | can read it | allowed-by-star |
| Amazonbot (Alexa, Rufus) | 200 | can read it | allowed-by-star |
The checks
- Answers AI agents like it answers people100%
- Readable without JavaScript100%
- robots.txt lets the crawlers in100%
- Facts in JSON-LD0%
- Content reachable, not buried60%
- Publishes a map of itself45%
- Plain structure82%
Sites with a similar score
Embed this score
Put the badge on data.gouv.fr — it links back here, and re-measures every time this index re-crawls.
<a href="https://shop.lumnika.com/ai-readiness/data.gouv.fr"><img src="https://shop.lumnika.com/ai-readiness/data.gouv.fr/badge.svg" alt="AI readability"></a>
Method. Six live HTTP requests plus robots.txt, llms.txt, sitemap.xml and security.txt, 12-second timeout each,
from a single vantage point. Blocking AI crawlers is a legitimate choice, not a failure: this page records what is true, not what should be.
Domain comes from a shop-platform seed (Shopify/WooCommerce/Magento/BigCommerce/Shopline fingerprint or JSON-LD price), not the Tranco traffic ranking. One page per domain — the homepage — is audited.
Part of the Agent Web Index, 47,298 domains measured, updated continuously.
Part of the Agent Web Index, 47,298 domains measured, updated continuously.