Metraipolis › archive › dc_45151aceef09

SkillFocus: Evolving Agent Skills via Capability Decomposition — arXiv:2609.34397

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.34397
License (as archived): CC0-1.0
Archived: 2026-10-07T20:57:44.511Z
SHA-256 of the archived text: 6ff629325eb73706ae2bbd8462319d73e7faff26b5b9a31c12aa73ae13adac24
Stable URL of this page: https://metraipolis.com/documento/dc_45151aceef09

How to cite

SkillFocus: Evolving Agent Skills via Capability Decomposition — arXiv:2609.34397 — arXiv (en), https://arxiv.org/abs/2609.34397 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_45151aceef09 (Metraipolis archive dc_45151aceef09, sha256 6ff629325eb73706ae2bbd8462319d73e7faff26b5b9a31c12aa73ae13adac24)

BibTeX

@misc{dc45151aceef09,
  title = {SkillFocus: Evolving Agent Skills via Capability Decomposition — arXiv:2609.34397},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.34397},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_45151aceef09; sha256 6ff629325eb73706ae2bbd8462319d73e7faff26b5b9a31c12aa73ae13adac24},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34397. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Agent skill evolution seeks to improve reusable procedural guidance for large language model (LLM) agents through iterative revision. Existing methods base each revision mainly on execution trajectories or feedback, leaving recurring behavioral requirements across tasks implicit and tying revision to the behavior of the current skill. We introduce SkillFocus, which decomposes recurring task requirements into a capability space that remains fixed as the skill evolves, separating what tasks require from how the current skill behaves. SkillFocus maps current task outcomes to this space to identify the capability that leaves the most tasks unresolved, then uses that capability to determine what to revise and which evidence to use. Across four benchmarks spanning heterogeneous tasks, SkillFocus achieves the best held-out accuracy on all four, outperforming the strongest competing result by 5.7 points on average while using 24\% fewer evolution tokens on average than the closest iterative baseline. Controlled studies further show that capabilities derived from recurring task requirements outperform task-semantic and execution-derived alternatives, while randomizing task--capability assignments reduces final accuracy by up to 20.2 points. Matching evidence to the selected capability increases candidate gain by 4.4 points under prioritized revision.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.34397 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T20:29:01.541Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).