编程Agent技能选不好？这篇论文用自回归解码一次搞定技能组合和顺序，在SkillsBench上直接涨了18到23个点，比传统检索还省token。
论文SkillComposer解决了编程Agent技能库选择瓶颈问题。传统方法将技能选择视为独立决策，而SkillComposer通过约束自回归解码器同时决策技能种类、数量和顺序。在SkillsBench上，结合GPT-5.2-Codex和Gemini-3-Pro-Preview，pass率分别提升23.1和18.2个百分点，超越top-3检索方法，并以更低提示词成本接近金标准技能上限。论文发表于arXiv:2606.32025。
Great paper on managing agent skills. Skill libraries keep growing, and picking the right skills ha...
Great paper on managing agent skills. Skill libraries keep growing, and picking the right skills has become a bottleneck for coding agents. The defaults are to expose the agent to the whole skill collection, or retrieve skills with embeddings and rerankers. Both treat the choice as independent picks. SkillComposer treats composition as one joint decision over which skills, how many, and in what order. A constrained autoregressive decoder over skill identifiers produces the full plan in a single pass, so dependencies between successive skills fall out naturally. On SkillsBench with GPT-5.2-Codex and Gemini-3-Pro-Preview, it lifts pass rate by +23.1 and +18.2pp over no-skill, beats top-3 retrieval, and matches the gold-skill upper bound at lower prompt-token cost. Paper: arxiv.org/abs/2606.32025 Learn to build effective AI agents in our academy: academy.dair.ai 💬 2 🔄 3 ❤️ 13 👀 1527 📊 6 ⚡