Test-case evaluation
Run representative inputs and score application outputs against defined criteria.
When to consider it
- When prompts or model-powered applications need repeatable behavior checks
Considerations
- Test cases and automated scores sample behavior; they do not prove universal correctness.
Connections
- In areaGovernance, Evaluation & Assurance
- Supports functionEvaluate system behavior
- Used byPromptfoo
Sources
Last reviewed 2026-09-30