Build or improve a reusable agent skill from a task description by acquiring practice data, testing a target model or coding agent, and refining instructions using held-out evaluation. Use when the user asks to learn a skill with Prompt2Skill or optimize a skill through execution feedback.
