Metraipolis › archive › dc_fd8ff88996ed

Skill++: Towards Lifelong Agents via Routing-Aware Skill Refinement — arXiv:2611.00544

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00544
License (as archived)
CC0-1.0
Archived
2026-11-03T10:41:56.036Z
SHA-256 of the archived text
e28749d0b79d02d013c0bb3fa3a98c849f409655ee936f30543be8a6fef00fde
Stable URL of this page
https://metraipolis.com/documento/dc_fd8ff88996ed

How to cite

Skill++: Towards Lifelong Agents via Routing-Aware Skill Refinement — arXiv:2611.00544 — arXiv (en), https://arxiv.org/abs/2611.00544 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fd8ff88996ed (Metraipolis archive dc_fd8ff88996ed, sha256 e28749d0b79d02d013c0bb3fa3a98c849f409655ee936f30543be8a6fef00fde)
BibTeX
@misc{dcfd8ff88996ed,
  title = {Skill++: Towards Lifelong Agents via Routing-Aware Skill Refinement — arXiv:2611.00544},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00544},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fd8ff88996ed; sha256 e28749d0b79d02d013c0bb3fa3a98c849f409655ee936f30543be8a6fef00fde},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00544. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large language model (LLM) agents extend their capabilities with skills, each with a name, a description, and a body. Under progressive disclosure, an LLM router selects skills by name and description alone and loads only the selected bodies. Skill libraries grow continually as users and agents add new skills. Because the router relies on this metadata, a similar new skill can displace an earlier one, breaking previously correct routing even though the earlier skill and the router are unchanged. This routing interference is aggravated for public skills, whose descriptions may not match their bodies. We formulate the problem as lifelong skill routing, a class-incremental continual learning problem. A solution must keep earlier skills selectable without suppressing new ones. Existing approaches fall short, since continually retraining the router risks catastrophic forgetting and rewriting methods either ignore similar skills or require per-skill feedback. We therefore propose Skill++, which rewrites only the name and description of each incoming skill, without router training or labeled data. The new description states what the body supports, how the skill differs from its closest predecessor, and what related work it excludes. As a library grows to 731 skills over ten sessions, Skill++ selects all required skills for 76.3% of queries, versus 36.4% under the original metadata. Across three benchmarks and five LLM routers, it outperforms all compared approaches, and its gains carry over to the Codex CLI agent harness.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2611.00544 — 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 2026-11-03T10:39:50.174Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).