Skip to contents

Guardrail Evaluation

guardrail_eval_result_class()
S7 class for guardrail evaluation results
guardrail_eval()
Evaluate a guardrail against a dataset
guardrail_metrics()
Compute guardrail evaluation metrics
guardrail_confusion()
Create a confusion matrix from guardrail evaluation
guardrail_compare()
Compare two guardrail evaluation results

Reports

guardrail_report()
Generate a guardrail evaluation report

Integration

benchmark_guardrail()
Benchmark a guardrail with positive and negative cases
benchmark_pipeline()
Benchmark a guardrail pipeline end-to-end

Vitals Interop

as_vitals_scorer()
Wrap a guardrail as a vitals-compatible scorer