wait-what

Invoke the wait-what skill explicitly and the assistant will re-pitch its previous message with added context. The rewrite uses ASD-STE100 Simplified Technical English and the ubiquitous language defined in your project's CONTEXT.md file.

ASD-STE100 is a controlled language standard that restricts vocabulary and sentence structure to make technical text clearer and less ambiguous. This Skill uses it to rewrite confusing AI responses into simpler, more precise wording.

No. The skill sets disable-model-invocation to true and allow_implicit_invocation to false, so it only runs when you explicitly invoke it. It will never fire on its own based on conversation context.

CONTEXT.md holds the project's ubiquitous language, the shared vocabulary agreed upon by the team. The skill references it so the simplified re-explanation uses the same domain terms as your project instead of generic wording.

Avoid it when the original response was factually wrong rather than unclear, since re-pitching only restates the same content more simply. For incorrect answers, ask for correction or additional sources instead of a simplified rewrite.