Metraipolis › archive › dc_01010698e89e

Robustness Analysis of Agentic AI to Inconsistent and Incomplete Tool Responses — arXiv:2608.22676

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.22676
License (as archived)
CC0-1.0
Archived
2026-10-18T07:41:34.049Z
SHA-256 of the archived text
4d36ea71c7ce5468960a77f2b9c93f12af206402643154bb9d39b097da85c19e
Stable URL of this page
https://metraipolis.com/documento/dc_01010698e89e

How to cite

Robustness Analysis of Agentic AI to Inconsistent and Incomplete Tool Responses — arXiv:2608.22676 — arXiv (en), https://arxiv.org/abs/2608.22676 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_01010698e89e (Metraipolis archive dc_01010698e89e, sha256 4d36ea71c7ce5468960a77f2b9c93f12af206402643154bb9d39b097da85c19e)
BibTeX
@misc{dc01010698e89e,
  title = {Robustness Analysis of Agentic AI to Inconsistent and Incomplete Tool Responses — arXiv:2608.22676},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.22676},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_01010698e89e; sha256 4d36ea71c7ce5468960a77f2b9c93f12af206402643154bb9d39b097da85c19e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.22676. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Tool-using agents increasingly rely on external tools to complete multi-step tasks, but tool returns can fail in different ways and require different recovery actions. Existing robustness studies often use uncertainty-based measures to detect when an agent becomes unreliable. These measures can reveal that something has gone wrong, but they do not directly identify the type of tool failure or the appropriate response. We address this limitation by analyzing tool failures at the moment a return enters the agent context. Our approach combines two complementary signals. The first compares the likelihood of the returned content under the tool schema and under the full trajectory prefix. The second measures the agent's probability distribution over its legal next actions. We evaluate the approach by injecting incomplete and inconsistent returns into a retail customer-service benchmark. The results show that likelihood-based signals clearly capture incomplete returns and some direct inconsistencies, while action-based signals reveal how strongly a failure changes the next decision. Some failures that are weak under likelihood signals can still redirect the agent toward state-changing actions. These findings show that tool failures can be recognized at the return boundary, but reliable diagnosis requires combining multiple signals.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-25 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.22676 — 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 2026-10-18T07:37:52.734Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).