Risk-next hotspots and how to act on them

SonarQube grades the code as it is. Risk Radar predicts where it breaks next.

Each file gets a 0-100 risk-next score fusing four signals that defect-prediction research (and our own data) agree on: severity-weighted issue density, churn, single-author ownership, and complexity. Files cluster their defects — the top of this list is where your next incident most likely originates.

How to use it:

1. Sort by risk, work top-down — don't spread effort evenly.
2. Read the "why this scored" factors; they point at the lever (tests, ownership, size).
3. Fix, re-scan, and watch the Outcome Ledger reconcile the prediction.

It's the decision tool for "where should the team spend this sprint's hardening budget?"