Metraipolis › archive › dc_2b30f07dfe17

Decision Models as Policy Gates for Coding Agents: Explicit Policy Boundaries Improve Detection and Reduce False Alarms — arXiv:2611.02155

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.02155
License (as archived)
CC0-1.0
Archived
2026-11-03T18:23:35.440Z
SHA-256 of the archived text
4152dcff75403d3164a6b20d682877d6962434d1863ac50fd258dc4347532578
Stable URL of this page
https://metraipolis.com/documento/dc_2b30f07dfe17

How to cite

Decision Models as Policy Gates for Coding Agents: Explicit Policy Boundaries Improve Detection and Reduce False Alarms — arXiv:2611.02155 — arXiv (en), https://arxiv.org/abs/2611.02155 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2b30f07dfe17 (Metraipolis archive dc_2b30f07dfe17, sha256 4152dcff75403d3164a6b20d682877d6962434d1863ac50fd258dc4347532578)
BibTeX
@misc{dc2b30f07dfe17,
  title = {Decision Models as Policy Gates for Coding Agents: Explicit Policy Boundaries Improve Detection and Reduce False Alarms — arXiv:2611.02155},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.02155},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2b30f07dfe17; sha256 4152dcff75403d3164a6b20d682877d6962434d1863ac50fd258dc4347532578},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02155. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Coding agents can destroy data, leak secrets and message the wrong person faster than a human can review. Could a decision model catch these risks at runtime? We evaluated 22 hosted decision models and two prompted general models as classifiers for a multi-rule policy engine. A synthetic dataset of 2,643 tool calls was labeled blind against 28 rules, with a sealed holdout. With a single shared preamble in both conditions, explicit boundaries written as examples and counterexamples raised recall by 5.5-26.1 percentage points on all 23 available models and reduced false positives on 21. Jev 1.14's recall rose from 89.3% to 98.2%, while unnecessary interruptions fell from 30.6% to 4.8% of 252 clean actions. With examples, its median latency was 120 ms at $0.46 per thousand requests. Strong models retained the recall gains when concrete names were replaced by generic descriptions; three largely retained them when the same boundaries were written as prose. On 16 new scenarios written after the initial studies, a follow-up improved recall on 22 of 23 models and reduced unnecessary interruptions on 21. We release the dataset, rule texts and harness.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Cryptography and Security (cs.CR); Software Engineering (cs.SE).
Pagina del lavoro: https://arxiv.org/abs/2611.02155 — 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 2026-11-03T18:10:07.957Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).