Metraipolis › archive › dc_06c35cdc217a
SkillLifeBench: Benchmarking Security Across the Lifecycle of Agent Skills — arXiv:2610.35215
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.35215
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-27T08:02:35.313Z
- SHA-256 of the archived text
8644a4a8c61da33a9485ff7b55fba96b5dde9e122222d3e1bbc577dd7100fe0d- Stable URL of this page
https://metraipolis.com/documento/dc_06c35cdc217a
How to cite
SkillLifeBench: Benchmarking Security Across the Lifecycle of Agent Skills — arXiv:2610.35215 — arXiv (en), https://arxiv.org/abs/2610.35215 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_06c35cdc217a (Metraipolis archive dc_06c35cdc217a, sha256 8644a4a8c61da33a9485ff7b55fba96b5dde9e122222d3e1bbc577dd7100fe0d)
BibTeX
@misc{dc06c35cdc217a,
title = {SkillLifeBench: Benchmarking Security Across the Lifecycle of Agent Skills — arXiv:2610.35215},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.35215},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_06c35cdc217a; sha256 8644a4a8c61da33a9485ff7b55fba96b5dde9e122222d3e1bbc577dd7100fe0d},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.35215. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Web agents reuse skills whose stored code, selected version, and loaded implementation can diverge across tasks and updates. SkillLifeBench links these states to task outcomes in 388 paired scenarios spanning implementation, interface, runtime, composition, and version changes. We evaluate 14 models on two paired decision tracks: 89 runtime tasks and 60 autonomous lifecycle tasks spanning 12 templates derived from introduced cases. Lifecycle tasks allow agents to inspect, select, and reconfigure skills while completing verifiable user requests. Joint success requires task completion without the registered attack goal. GPT-5.6-Sol and Sonnet 4.6 both achieve 82.0% on runtime tasks, but reach 96.7% and 56.7% on lifecycle tasks. Across lifecycle evaluations, challenged task success is 77.5%, while joint success is 65.2%; 103 runs complete the task and the attack goal. Adding execution provenance in a matched recovery study raises utility joint success from 49.35% with logs and metadata to 65.58%. A deterministic execution-consistency rule reaches 64.94% with higher benign utility, while full restoration reaches 69.48%. Lifecycle tasks distinguish models with similar runtime scores, and execution evidence improves recovery. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.35215 — 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 2026-10-27T07:55:30.974Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).