using-superpowers

To enforce pre-response skill invocation, you need a system that checks for applicable skills and loads them deterministically before generating any user-facing reply. This prevents ungrounded outputs and aligns conversations with user instructions.

Automatic skill invocation is a pre-response check mechanism that detects applicable skills across all conversations. It works by loading and executing the appropriate skills before the assistant generates any text, preserving user instructions as the highest priority.

You set up global guardrails by enabling a skill tool that applies across all conversations and tasks. This mechanism prevents responses that ignore safety prompts or bypass applicable skills, ensuring policy enforcement before any reply.

Yes, you can use priority handling to ensure user instructions override applied skills. The system respects user instructions above other content and applies skills accordingly, satisfying the requirement to preserve user directives as the highest priority.

Yes, pre-response skill invocation works during clarification questions. It ensures skills are checked and engaged before any user-facing reply, including clarifications, preventing outputs that bypass safety prompts or ignore applicable skills.

You should avoid deterministic skill loading when a conversation requires an immediate, unmediated response without pre-response checks. Bypassing this mechanism risks generating ungrounded or unsafe outputs that ignore established policy enforcement and safety guardrails.