using-superpowers

To enforce skill discovery before an agent responds, you mandate explicit Skill tool invocation during initial conversation handling and plan phases. This ensures relevant skills are checked and loaded prior to producing any assistant output or asking clarifying questions.

Selecting an agent workflow requires applying priority rules to choose between checklist-driven-development, subagent-driven-development, and executing-plans workflows. You must order process skills before implementation skills to determine the correct workflow.

Your assistant takes premature actions because it lacks a mandated skill check before responses. By enforcing strict adherence to loaded skill instructions and explicit Skill tool invocation, you prevent missed workflows and inconsistent procedures across conversations.

Yes, enforcing skill invocation works during brainstorming, debugging, and implementation workflows. It applies whenever tasks might benefit from a skill check, requiring explicit Skill tool invocation before any clarifying questions or actions are taken.

Yes, you can force an agent to run brainstorming first when asked to build a feature. The agent runs the brainstorming skill, then selects the appropriate implementation skill and workflow based on established priority rules.