ax-refine
Use this skill when writing or reviewing Ax bestOfN/refine code, reward functions, thresholds, native sample selection, serial attempts, generated advice, and attempt diagnostics.
No security review available
No security review has been published for this skill
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.