Test Confidence Report — selfaudit (Developer Report)

Confidence Score

14.3%

Tests providing reliable signal — run correctly, test real behavior, cover critical code

Risk Level

High Risk

Overall assessment of release readiness based on coverage alignment

Displacement Score

85.7%

Tests passing CI that do not protect meaningful behavior or critical code paths

What This Means for Your Next Release

This codebase carries high release risk. At 14% confidence, the majority of tests passing CI are not protecting meaningful behavior. The highest concern is that Agents, Logging — your most critical code paths — are not adequately covered by integration or end-to-end tests. If these areas fail in production, your current tests will not catch it. 86% of tests either validate implementation details instead of behavior, or cover low-risk code — creating a false sense of safety in CI. There are 2 P0 items that should be resolved before the next release. Shipping without addressing P0 gaps means relying on a test suite that does not protect the code that matters most.

Summary

86% of tests validate real behavior — but 86% are misaligned to low-risk code paths. Your tests run correctly and test real behavior, but they are covering the wrong areas. A passing CI pipeline does not mean your critical code paths are protected.

Top finding: Tier 1 critical path infrastructure has zero integration or e2e coverage, leaving core system behavior entirely unvalidated beyond isolated unit boundaries.

↑ Top