The audit report
The artifact that survives the budget meeting.
Every engagement ends in a document a CTO can defend: one health score, 40 scored dimensions, and every finding priced by what it costs to fix.
Explore the real sample · halyard-api, score 6.6
What’s inside
Eight pages. Zero filler.
01Cover with a live fixYour actual highest-severity deterministic finding, before and after: a real bug, fixed and verified, free.
02Executive summaryScore, scope, severity breakdown, and the fix-economics split, readable in ten seconds.
03Clustered scorecardForty dimensions rolled into seven families, each with its weakest and strongest member.
04Full 40-dimension detailEvery dimension scored 0–10, sorted worst-first, on a single page.
05Findings by fix costEverything free to fix leads; cataloged and structural follow. Each carries evidence, a score delta, and a disclosed top-30-of-N cap.
06Compliance & governanceNine frameworks, SOC 2 through EU AI Act, with real control citations backed by real findings, plus per-control detail for the worst scorer.
07Framework control detailThe worst-scoring framework, article by article. Hit rows expand into the findings that cite them; clean rows say so.
08Methodology & validationHow the numbers are made: static AST analysis, three-tier classification, re-parse verification.
✓ every number maps to a real engine field
✓ validated: ast-grep re-parse + lint
✓ print, web, and phone renderers, one payload