behavioral-modes

To switch AI behavior modes, ask your AI to enter a specific mode using commands like '/brainstorm' or '/debug', which loads task instructions and enforces mode-specific constraints to optimize decision-making for that context.

The best way to structure AI responses for debugging is to use a dedicated DEBUG mode, which enforces systematic problem solving, root-cause analysis, and reproducible fixes rather than allowing generalized, unstructured answers.

Yes, you can use adaptive AI modes for both tasks. BRAINSTORM mode frames early ideation and explores alternatives, while IMPLEMENT mode focuses on writing code, building features, and executing plans according to quality standards.

Mode-switching improves review and deployment by loading distinct REVIEW and SHIP modes. REVIEW mode enforces thorough, constructive architectural analysis with clear rationale, while SHIP mode enforces production readiness and stability constraints.

No dependencies are required to enforce mode-specific constraints. The skill operates independently, loading its task instructions directly from its own frontmatter and body to adapt AI behavior across planning, verifying, and teaching contexts.

You should avoid using a single AI mode for an entire workflow when tasks span multiple contexts like ideation and deployment. Switching modes ensures task-appropriate constraints, preventing generalized outputs that lack the focus needed for specific phases.