Eval Tracking
Skill for Supabase-backed evaluation result tracking.
Overview
Track evaluations with:
eval_runs- Evaluation run metadataeval_cases- Individual test caseseval_scores- Metric scores per case
Use When
This skill is automatically invoked when:
- Storing evaluation results
- Building eval dashboards
- Tracking regression over time
- Comparing run results
Available Scripts
| Script | Description |
|---|---|
scripts/setup-tracking.sh | Run Supabase migration |
Available Templates
| Template | Description |
|---|---|
templates/schema.sql | Supabase tables and RLS |
templates/queries.sql | Dashboard queries |
