# Changelog

## 0.2.0 - 2026-09-28

First public release under MIT.

- `scoreDimension()`, `scoreProduct()` and `scoreSuite()` apply a
  five-dimension readiness rubric with hard caps and six bands.
- The lowest open cap on a dimension wins; an unknown cap id throws.
- A score above 70 with no evidence is held at 70 and reported as NOT
  ASSESSED.
- The overall score is the weakest dimension; the average is reported
  alongside and labelled as such.
- Suite-wide caps apply before the tier-weighted average.
- A cap can cite the evidence that establishes it; an uncited cap is still
  applied and the report says so.
- `productReport()`, `suiteReport()`, `formatProduct()` and `formatSuite()`
  produce an evidence-record report.
- Worked example in `examples/cap-by-worst-defect.ts`.