Can an agent find canonical capabilities and constraints?
Public evidence assessment · v1.0
Is your platform ready
for its next user?
Evaluate whether AI agents can discover, understand, access, and safely operate your B2B product—not just browse it.
01 / METHOD
Readiness is more than
having an API.
Agentic users need interfaces they can discover, permissions they cannot quietly exceed, outcomes they can verify, and failures they can recover from.
Are actions, schemas, costs, and errors machine-legible?
Can it authenticate, act, retry, and observe safely?
Can humans constrain, audit, reverse, and stop it?
02 / CHECKLIST
The agent-ready
platform checklist.
Thirty practical controls across eight dimensions. Publicly detectable items feed the score; internal controls require evidence from your team.