// free tool
Generate the whole agent pack.
One scan, every machine-readable file agents look for: llms.txt, agents.txt, JSON-LD, an A2A agent card and a UCP commerce manifest.
Questions
What's in the pack?
Five files: llms.txt (a curated site map for assistants), agents.txt (which AI agents may use the site), a JSON-LD block for the scanned page, a /.well-known/agent-card.json A2A card, and a /.well-known/ucp.json commerce manifest.
Can I publish these as-is?
The llms.txt and JSON-LD, yes — they're built from what's actually on your site. The agent card and UCP manifest contain placeholder endpoints, and we mark them as stubs. Publishing a manifest that points at endpoints you haven't built is worse than publishing none, because it fails for everyone who tries it.
Which of these are real standards?
Honestly: they vary. JSON-LD is well-established. llms.txt is a widely-discussed convention with uneven adoption. A2A and UCP are young and still moving. agents.txt is the least settled of the lot — it states intent, while robots.txt remains what crawlers actually enforce.
Is anything sent to an AI model?
No. Every file is generated from deterministic templates using what we fetched from your site. Same input, same output, no per-scan cost, and nothing from your pages ever reaches a prompt.
How is this different from the Agent Readiness Checker?
That tool tells you what's missing and why. This one skips the diagnosis and just hands you the files. If you want to know what's broken first, start there.
// ready_when_you_are
Let's build your agentic advantage
Tell us what should be automated, connected or sold. We'll map the fastest path from idea to a production system.