Link to view in the Playground
An ontology modeling the full lifecycle of AI agent and skill evaluation — from test case design and benchmark sets through execution, failure analysis, and remediation. It gives teams a shared vocabulary for systematic quality work, formalizing the relationships between what gets tested, how success is defined, and what happens when something breaks.
