using-superpowers

To ensure correct skill invocation, you must enforce a mandatory pre-check behavior that checks and invokes applicable skills before taking action or responding to any chat interaction. This prevents wasted effort and incorrect workflows by verifying skill relevance first.

Priority rules for user instructions dictate that user instructions always remain the highest priority, while superpowers skills can override default behavior. This hierarchy clarifies how to handle process-first versus implementation-second situations during workflow execution.

Yes, skill invocation works across Claude Code, Copilot CLI, and Gemini CLI. The skill provides specific platform mapping and access instructions to reliably activate the correct instructions and prompt routing across these environments.

You should invoke skill checking immediately when starting a conversation, before asking clarifying questions or producing any response. This mandatory pre-check applies whenever tool-driven skill execution might be relevant, even if the scenario seems small or unclear.

Your agent skips relevant skills because it lacks a mandatory Skill-tool invocation policy. Forcing the agent to check whether a relevant skill applies before responding or taking action ensures workflow-aligned behavior and prevents incorrect execution paths.