Skills

Use skills to give AI workflows repeatable instructions, patterns, and domain-specific guidance.

Skills are reusable packets of guidance that help AI behave more consistently in a specific kind of task. They are useful when the same planning pattern, documentation standard, or operating rule should be applied repeatedly.

What a skill usually contains

A skill may define:

• the type of task it is for
• the structure of a good output
• rules or constraints the AI should follow
• references or examples that improve consistency

Why skills matter

Without reusable guidance, AI outputs vary too much from one run to the next. Skills reduce that variation and make the workflow easier to scale across a team.

Keep skills narrow

The best skills are focused. A skill for variance summaries is easier to trust than one giant skill that tries to cover every planning task at once.