domains / families / agent-tooling / evals-bootstrap

evals-bootstrap

separate

Scaffold a first eval suite for an agent: mine real failures into cases, write behavioural checks over traces, and generate the runner. Use when the user wants evals, regression tests for an agent, a golden set, or asks how to know a prompt/model change did not break things. Do NOT use for a single task's done-check (goal-test) or for auditing context (context-audit).

When to use

• scaffold a first eval suite
• mine real failures into cases
• write behavioural checks over traces
• generate the runner
• regression tests for an agent
• create a golden set
• verify prompt or model changes did not break things

When not to use

• single task's done-check
• goal-test
• auditing context
• context-audit

Skill page

Source