Agent-Readiness Audit

API-sampled

Who AI recommends

Category prompts for the top 20 index sites, plus any Pro site, are sent to the model APIs that have a key. Each prompt is sampled more than once so the list can disagree with itself. This is not evidence that a recommendation caused a win or a loss.

Back to the index

No sample is stored yet. The Monday pass tracks the top 20 scored sites for four weeks. An engine is skipped when its API key is missing.