Probe a host
Reads https://<host>/.well-known/ai-catalog.json in your browser — no server, no logging.
That Well-Known URI is the spec's primary discovery path. ARD also lets a publisher point to a catalog elsewhere — this explorer doesn't follow these:
- robots.txt — an
Agentmap:directive, e.g.Agentmap: https://example.com/catalog.json - HTML link tag —
<link rel="ai-catalog" href="…">in the page<head>
Catalog advertised that way (and not at the well-known path)? Paste its full URL in the box above, or use Paste JSON.
Try
- Reach
- Fetch
- Parse
- Validate
- Verdict