Artificial Intelligence · 06.08.2026, 00:53 UTC
Microsoft’s SkillOpt Shows Optimized Agent Skill Artifacts Transfer Across Model Scales and Between Codex and Claude Code Harnesses
| Schweregrad | info |
|---|---|
| Kategorie | Artificial Intelligence |
| Quelle | MarkTechPost ↗ |
| Veröffentlicht | 06.08.2026 UTC |
Sicherheitsmeldung mit Schweregrad noch nicht bewertet. Technische Details im Tab „Originaltext“; empfohlene Schritte in der Checkliste.
SkillOpt is a text-space optimizer developed by a team of researchers from Microsoft, Shanghai Jiao Tong University, Tongji University, and Fudan University.
SkillOpt trains a single natural-language skill document while the target model stays frozen. An optimizer model reads scored rollouts and proposes bounded add/delete/replace edits. A held-out selection split accepts an edit only when the score strictly improves. The exported artifact is one file, best_skill.md.
The transfer tables report three columns. Baseline is the target’s no-skill score. Direct is SkillOpt trained in-domain on that exact target. Transferred applies a skill trained elsewhere, with no further optimization.
The useful comparison is not transferred versus direct. It is how much of the in-domain gain survives the move.
Cross-model transfer: within-family, mixed retention
Skills were trained on GPT-5.4 and deployed on smaller variants.
SpreadsheetBenchGPT-5.4-mini36.147.545.5+9.482%SpreadsheetBenchGPT-5.4-nano23.542.526.5+3.016%LiveMathGPT-5.4-mini14.732.819.2+4.525%LiveMathGPT-5.4-nano23.227.228.8+5.6140%
Two rows deserve attention. SpreadsheetBench on GPT-5.4-mini keeps 82% of the in-domain gain. That is close to free reuse. The LiveMath row on GPT-5.4-nano is stranger: the transferred skill scores 28.8 against an in-domain SkillOpt result of 27.2. The paper reads this as evidence that some learned procedures are target-model agnostic.
The GPT-5.4-nano SpreadsheetBench row is the weak one at 16%. Retention is not uniform, and the paper does not claim it is. Its stated bound is narrower: no …