Trust asset

What a reproducible audit finding looks like.

A useful finding is not just a paragraph in a PDF. It gives the team a concrete behavior to reproduce, a reason to care, and a way to verify the fix.

Finding qualityRerunnable
Before fix

The reproducer fails or demonstrates the issue.

After fix

The same evidence verifies the patch or shows what remains.

Structure

Proof findings are built for engineering action.

1. Behavior mismatch

What the system should do, what it actually does, and which requirement or invariant is affected.

2. Reproducer

Minimal steps, test case, input, or counterexample that demonstrates the behavior.

3. Impact

Severity, affected surface, blast radius, exploitability, customer impact, and regression likelihood.

4. Evidence

Tests, traces, coverage, MC/DC, formal artifacts, or manual review confirmation.

5. Remediation

Fix guidance aimed at the bug class, not just the one observed instance.

6. Retest

Patch verification using the same evidence corpus whenever possible.

Why it matters

Reproducibility changes the economics of audit work.

A static finding loses value as the codebase moves. A reproducible finding can become a regression test, release gate, patch-verification artifact, customer evidence, or disclosure-safe case study.