architecture

Automated layer discovery scans your codebase directories to identify natural layers and existing cross-reference patterns. It generates an editable YAML configuration file defining allowed dependencies, which a post-edit hook uses to validate changes and prevent architectural drift.

You can automatically discover codebase layers by running an architecture initialization command that scans top-level directories. This process identifies existing cross-directory dependency patterns and outputs a human-readable YAML boundary configuration file for your team to review and adjust.

Yes, the generated YAML boundary configuration is fully editable. After the automatic layer discovery process identifies your codebase's natural layers and existing cross-references, your team can freely adjust the allowed dependency rules within the YAML file to match your architectural requirements.

The best way to warn developers is integrating a post-edit hook that validates changes against defined architecture boundaries in real-time. This immediate validation warns developers when an edit introduces an unapproved cross-layer reference, preventing accidental architectural drift.

Yes, dependency boundary enforcement works for projects with distinct directories like rules, scripts, skills, and agents. It detects existing cross-directory dependency patterns, generates a boundary config, and provides warnings when edits introduce unapproved cross-layer references between these layers.