Code health & tech-debt checks
Beyond correctness, QualityMax reports code health: the tech-debt scan that ships with every pipeline run. Its findings appear as a dedicated GitHub check run named QualityMax Code Health, with line-level annotations on the exact code that accumulated debt — the same shape as a human reviewer’s inline comments.
What the check reports
Section titled “What the check reports”- Maintainability findings — complexity hotspots, god-functions, deep nesting
- Duplication and dead-code signals
- Severity-graded findings, each attached to a file and line
Each PR gets a check title summarizing the verdict and a body listing the findings; annotations are batched to respect GitHub’s per-request limits (50 per request) and mapped from internal severities to GitHub annotation levels (warning, failure, notice).
Verdict to conclusion
Section titled “Verdict to conclusion”| Scan verdict | Check conclusion |
|---|---|
| Pass | success |
| Findings | action_required / failure by severity |
Like Gate Delta, code-health findings inform rather than ambush: the check is visible on every PR, annotated to the line, and stable enough to trend across commits.
Using it well
Section titled “Using it well”- Treat annotations as a debt ledger: the same hotspot flagged across PRs is a refactor candidate, not a per-PR nuisance.
- Pair with the complexity and dead-code skills to rank what to pay down first.
- The check is per-PR, so debt introduced by the change is distinguishable from pre-existing debt it merely touched.