Microsoft and three Chinese universities have developed SkillOpt, a method that optimizes instruction documents for AI agents using principles from traditional model training. A simple Markdown file is enough to boost GPT-5.5 by about 23 points on procedural tasks, and the same file transfers across models and agent environments like Codex and Claude Code.

Microsoft's SkillOpt brings deep-learning discipline to AI agent skills, replacing manual prompt tweaking with mathematically validated text optimization.

After building 50+ AI systems, here is what we know about optimizing AI agent skills. SkillOpt is a...