using-superpowers

Skill-driven response gating ensures the AI checks for applicable skills before generating a response. By invoking the Skill tool at the start of a conversation, it loads the most relevant Skill to prevent misaligned answers and enforce consistent behavior.

The best way to load relevant skills is to invoke the Skill tool before replying. This mechanism guides how to search for, load, and execute the most appropriate Skill to handle clarifications and complex tasks safely.

Skill-driven response gating handles complex tasks by guiding users to search, load, and execute applicable Skills. This ensures that task execution and clarifications follow a safe, rule-driven approach based on the loaded Skill's frontmatter.

Misaligned answers occur when the AI replies without checking for applicable skills. Invoking the Skill tool before any response ensures safe, rule-driven behavior by loading the most relevant frontmatter and executing the proper workflow.

The primary limitation is that this approach enforces a strict workflow, requiring a Skill tool invocation before any reply. This gating mechanism ensures safe behavior but may add processing overhead to simple clarifications across all conversations.