using-superpowers

To enforce tool invocation before responding, apply a skills-first protocol that mandates AI agents check and utilize available skills before generating any output, preventing unassisted actions and ensuring structured task execution.

A skills-first protocol is a workflow enforcement rule requiring AI agents to invoke relevant skills for information retrieval or task completion before providing any answer, overriding any tendency to prematurely respond or self-assist.

To stop AI agents from skipping tool checks, implement a mandatory skill invocation rule that applies to all conversational turns, overriding rationalizations to self-assist and requiring strict adherence to the defined process.

Yes, the mandatory skill invocation protocol applies to all conversational turns and task execution scenarios, ensuring AI agents utilize available tools for information retrieval or process guidance before generating any output.

You need to enforce skill invocation before generating an AI response whenever a user asks a question that could potentially be answered by a skill, requiring the AI to gather information or perform an action first.