AI 精选动态
智能评分 65
SkillComposer论文：通过联合解码提升Agent技能组合
AI 推荐理由
与独立检索技能不同，该方法通过自回归解码联合考虑技能顺序，值得深入阅读以了解实现细节。核心解读
SkillComposer提出将技能组合视为联合决策，使用约束自回归解码器一次生成完整计划。在SkillsBench上使用GPT-5.2-Codex和Gemini-3-Pro-Preview，pass率较无技能提升+23.1和+18.2个百分点，超过top-3检索，匹配gold-skill上界且token成本更低。
全文
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: https://t.co/ovbQf07Mmk
Learn to build effective AI agents in our academy: https://t.co/1e8RZKs4uX
