using-superpowers

Enforcing mandatory skill invocation requires a strict protocol that guides the AI through a decision tree to check for relevant skills before taking any action. This prevents unguided or inefficient task execution by requiring the AI to verify applicability.

Copilot agentic workflows often skip tools due to common rationalizations for bypassing skill checks. Implementing a skills-first protocol identifies these red flags and forces the AI to invoke relevant skills even when the probability of applicability is low.

A skills-first protocol is a decision process that ensures AI agents always check for and invoke relevant skills before responding. It solves the problem of AI agents acting without leveraging available specialized tools or knowledge bases.

To determine relevance before responding, the AI must evaluate if a skill like 'csharp-docs' or 'dotnet-upgrade' applies and invoke it if there is even a small chance. This structured workflow guidance mandates checking before asking clarifying questions.

You need to enforce skill selection when AI agents are acting without leveraging available specialized tools or knowledge bases. It is required before any AI response or action to prevent unguided task execution and ensure strict adherence to the invocation protocol.

Yes, you can use a decision tree to provide structured workflow guidance for when and how to invoke skills. This approach enforces the rule to use skills even with a low probability of applicability and lists common rationalizations for skipping checks.