en un clic
eval-engineer
eval-engineer contient 8 skills collectées depuis Galileo-Agent-Labs, avec une couverture métier par dépôt et des pages de détail sur le site.
Skills dans ce dépôt
Use when Galileo evidence is available and the user asks why a trace, session, log stream, experiment, metric, or AI app behavior failed, regressed, or became unsafe.
Use when a user is unsure which Eval Engineer command to run for AI agents/RAG apps, needs onboarding/status for a .galileo workspace, or asks where to start.
Use when the user asks to set up Eval Engineer, check .galileo readiness, create workspace scaffolding, or configure editable files, verification commands, app type, or evidence paths.
Use when the user asks for an AI app audit, launch readiness review, safety/security review, OWASP agentic risk check, metric coverage review, or production RCA gap review.
Use when the user asks to make an AI app cheaper or faster, reduce tokens, latency, model/tool/retrieval/rerank/self-check/retry/evaluator cost, or compare cost before/after.
Use when the user asks to turn a failure into an eval, create/review/accept/reject dataset cases, or convert Galileo traces, metric gaps, or production examples into cases.
Use when a user asks to fetch Galileo evidence, provides Galileo URLs or IDs, says "fetch this Galileo link", or needs traces, sessions, experiments, or log streams saved locally.
Use when the user asks if an AI app is measured correctly, needs Galileo metrics, expected-output contracts, metric profiles, eval gates, or measurement before optimizing.