frugal — save tokens, never save effort

The one rule decides every judgment call: the quality gate must not move. A "cheaper" run that fails the task, retries twice, or ships unverified work costs more than it saved. The KPI is successful task token cost — never raw token minimization.

When NOT to use

• The task is short and context is healthy — overhead exceeds savings.
• The user asked for exhaustive output deliberately.
• A "saving" would remove verification steps — that's the effort this skill exists to protect.

The techniques

Grouped by where the tokens go. Each is a habit to apply when it fits, not a checklist to run ritually.