prompt-token-efficiency

スキルの内容

A general skill stored in the gh-aw repository for rewriting prompts so they use fewer tokens while keeping their meaning. It sets goals of fewer tokens, clear wording and little ambiguity, then gives rules: keep only what the task needs, drop pleasantries and repetition, compress prose into short directives, and run checks for ambiguity and lost meaning after the rewrite. It applies to any prompt written for a language model and does not involve GitHub.

できること

詳細