dark-code-audit

To identify comprehension gaps in a codebase, you can generate a darkness report that analyzes authorship signals, decision trails, behavioral contracts, and dependency relationships to produce per-module scores and a heat map.

You can audit AI-generated code risk by scanning the repository to produce a darkness report that quantifies missing human comprehension and outputs prioritized TODOs for remediation.

Yes, the darkness scoring process performs module discovery across languages and boundaries, evaluating authorship signals and dependency awareness to quantify comprehension gaps per module.

The best way to visualize codebase health is to generate an aggregate heat map that highlights per-module darkness scores, providing a clear view of where code understanding is missing across the repository.

Yes, the darkness report outputs prioritized TODOs for remediation alongside the heat map, making it suitable for onboarding new repositories by highlighting exactly where code comprehension needs improvement.

You should use a darkness report when you need to quantify human comprehension gaps rather than test execution coverage, such as when auditing AI-generated code risk or performing periodic health checks across large codebases.