bkit-rules

Agent auto-trigger rules proactively select appropriate assistants for code review, design validation, and gap analysis based on user intent. They detect project maturity levels to apply corresponding agent behaviors, optimizing code quality outcomes without manual selection.

You can use level-based guidance to detect project maturity and apply corresponding agent behaviors for code tasks. This level-aware check mechanism optimizes outcomes by tailoring design-first decisions and quality enforcement to the specific development context.

The best way to maintain consistency is applying a formalized PDCA rule set with level-aware checks and guardrails. This enforces planning, design, analysis, and reporting steps while promoting DRY, SRP, and testable patterns across all AI-native coding tasks.

Automated PDCA rules may not suit rapid prototyping or experimental coding where formal planning, design, and reporting steps create unnecessary overhead. The enforcement of structured workflows is best suited for production-level feature development, bug fixes, and refactoring.