AI Agents / Agent evaluation and observability
AI Agents: Task success rubric
Score completion against observable artifacts and required constraints.
Environment and inputs: Framework-independent agent evaluation; OpenTelemetry illustrates trace correlation.
When to use it
Choose an observable task outcome before optimizing a model or prompt. Keep inputs and scoring rules stable enough to compare candidates, and inspect critical failures individually rather than hiding them inside an average quality score.
Check the result
Have two reviewers apply the rubric to ambiguous examples and reconcile disagreements.
Save and compare results
Save the test input, configuration and result. Fix the failing boundary, repeat the same test and compare it with the saved working configuration.
Scope and limits
These exercises have not been tested on your systems. Check feature support and configuration names against the installed release and linked documentation.