codebase-audit-grounding

To quantify your codebase before writing specifications, you need a codebase audit that measures current metrics like test files, error boundaries, and dependency counts. This generates a measured current-state report to anchor your specs.

Codebase metrics needed for specification-driven development include counts of test files, aria markers, error boundaries, localStorage usage, and dependencies. These measured data points eliminate speculation by providing a quantified snapshot of the actual repository state.

You audit your repository for test files and dependency counts by executing grep and find commands across the file system and reading package manifests. This produces a concise measured report of your codebase's current state for specification grounding.

Yes, a codebase audit requires file system access to execute grep and find commands and read package manifests. This access is necessary to collect measured data on test, accessibility, error handling, storage, and dependency metrics for your specification.

You should perform a codebase audit for an architectural review before writing new specs or when assessing the impact of changes in a repository. It provides a quantified current-state report to ground architectural decisions in actual metrics rather than assumptions.