Metraipolis › archive › dc_ab4d73b8b2de
CoRedSkill: Stealthy Skill Composition Attacks via AST-Guided Injection — arXiv:2610.20049
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.20049
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T20:02:53.152Z
- SHA-256 of the archived text
f17664d2f04a33515e552c0224a20e5a957f29f96be87660f3ebde7335956aaf- Stable URL of this page
https://metraipolis.com/documento/dc_ab4d73b8b2de
How to cite
CoRedSkill: Stealthy Skill Composition Attacks via AST-Guided Injection — arXiv:2610.20049 — arXiv (en), https://arxiv.org/abs/2610.20049 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ab4d73b8b2de (Metraipolis archive dc_ab4d73b8b2de, sha256 f17664d2f04a33515e552c0224a20e5a957f29f96be87660f3ebde7335956aaf)
BibTeX
@misc{dcab4d73b8b2de,
title = {CoRedSkill: Stealthy Skill Composition Attacks via AST-Guided Injection — arXiv:2610.20049},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.20049},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ab4d73b8b2de; sha256 f17664d2f04a33515e552c0224a20e5a957f29f96be87660f3ebde7335956aaf},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.20049. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Agents extend their capabilities through reusable skills that package instructions, executable code, and resources. Composing these skills enables complex task workflows, but their dependencies can also connect malicious operations across skills, making their combined intent difficult to identify through individual-package auditing. We introduce CoRedSkill, a compositional red-teaming method that forms stealthy attacks through cooperating skills while preserving their original utility. CoRedSkill decomposes an existing workflow into interdependent skills and embeds attack operations in task functions through abstract syntax tree (AST) rewriting, with skill instructions guiding the agent to autonomously select and invoke the relevant skills, and workflow dependencies and intermediate artifacts connecting the injected operations during task execution. We construct a benchmark of 150 cases covering 50 tasks and three attack objectives to evaluate attack success, task completion, and audit detection. In the main evaluation, CoRedSkill achieves 98.0% attack success and 95.3% task success with a 3.3% detection rate, and a joint success rate of 90.7% (attack success, task completion, and audit evasion in the same case) compared with 6.7--42.7% for the four baselines. Skill compositions can therefore complete legitimate tasks while carrying out unauthorized operations that evade auditing. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.20049 — 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 2026-10-15T19:35:48.395Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).