Metraipolis › archive › dc_fd73654d424c
RedEvoAgent: Automatic Red-Teaming Agent with Experience-Driven Skill Evolution — arXiv:2608.27439
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.27439
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T14:39:25.098Z
- SHA-256 of the archived text
6606c185e1adf91af44f24643b67056f9bee00d1feafa7c8dbf675d70ea391bb- Stable URL of this page
https://metraipolis.com/documento/dc_fd73654d424c
How to cite
RedEvoAgent: Automatic Red-Teaming Agent with Experience-Driven Skill Evolution — arXiv:2608.27439 — arXiv (en), https://arxiv.org/abs/2608.27439 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fd73654d424c (Metraipolis archive dc_fd73654d424c, sha256 6606c185e1adf91af44f24643b67056f9bee00d1feafa7c8dbf675d70ea391bb)
BibTeX
@misc{dcfd73654d424c,
title = {RedEvoAgent: Automatic Red-Teaming Agent with Experience-Driven Skill Evolution — arXiv:2608.27439},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.27439},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fd73654d424c; sha256 6606c185e1adf91af44f24643b67056f9bee00d1feafa7c8dbf675d70ea391bb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.27439. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
LLM-based agents are increasingly deployed in product-level execution harnesses, where jailbreaks can trigger harmful tool use and persistent state changes, creating greater risks than unsafe text generation alone. Existing automatic red-teaming methods often rely on fixed attacks, while recent agentic attackers coordinate multiple jailbreak tools and show stronger potential through trajectory-based retrieval. However, such retrieval can reuse misleading experiences due to retrieval bias and unclear tool credit, and full trajectories add context overhead while reducing interpretability. We propose RedEvoAgent, a black-box red-teaming agent that distills cross-case attack trajectories into a concise, human-readable attack skill. The attack skill adaptively evolves through tool-effectiveness profiling and Deciding-Tool Attribution for skill updates, and a validation ratchet that retains only updates improving validation performance. Experiments on multiple benchmarks, target models, and target execution harnesses show that RedEvoAgent outperforms fixed and agentic baselines, improves tool efficiency, and transfers across attacker models and target execution harnesses. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-28 in cs.AI). Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.27439 — 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 2026-10-17T13:16:53.727Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).