Metraipolis › archive › dc_967fa6904095

NOMOS: Compiling Written Policies into Statically Verified Tool-Call Gates for LLM Agents — arXiv:2610.11030

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.11030
License (as archived)
CC0-1.0
Archived
2026-10-09T05:56:34.451Z
SHA-256 of the archived text
73fe154c39705039bbd68bed91663ef7b5473df1756ea84bed2966fa55118b56
Stable URL of this page
https://metraipolis.com/documento/dc_967fa6904095

How to cite

NOMOS: Compiling Written Policies into Statically Verified Tool-Call Gates for LLM Agents — arXiv:2610.11030 — arXiv (en), https://arxiv.org/abs/2610.11030 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_967fa6904095 (Metraipolis archive dc_967fa6904095, sha256 73fe154c39705039bbd68bed91663ef7b5473df1756ea84bed2966fa55118b56)
BibTeX
@misc{dc967fa6904095,
  title = {NOMOS: Compiling Written Policies into Statically Verified Tool-Call Gates for LLM Agents — arXiv:2610.11030},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11030},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_967fa6904095; sha256 73fe154c39705039bbd68bed91663ef7b5473df1756ea84bed2966fa55118b56},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11030. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Tool-using LLM agents violate the policies they are deployed to enforce, often silently. Prior defenses hand-write rules, query an LLM verifier per action, or compile policies through heavyweight formal machinery. Naive compilation fails: extracted rules block the tool satisfying their own precondition, or read arguments their tool lacks. NOMOS, a four-pass compiler, turns a natural-language policy into a deterministic tool-call gate; static verification with tool-schema-level checks alone (no prover, solver, or LLM) repairs or rejects 37% (airline) and 13% (retail) of candidates, without which most shipped rules are inoperable. Replaying compiled rules over undefended transcripts flags bindings that refuse legitimate work (a development binding refused 95.9% of task-passing calls); no evaluation binding is flagged. On $\tau^2$-bench the gate cuts violations of reference-encoded clauses among state-changing calls from 66.3% to 2.6% (airline) and 30.8% to 6.9% (retail), raising airline task success significantly for $2 \le k \le 4$; a 26B on-premise compilation is not significantly worse than hand-written or frontier-compiled rules. Unlike AgentDojo's shipped defenses, it reaches a zero attack success rate (ASR) on banking, where nine attack families collapse onto three structural rules. On the other three suites its ASR is at most 3.6%, from goals with no tool call to govern and one write admitted by a binding weaker than its clause; a second agent model, Llama-3.3-70B, reproduces the effect on both benchmarks. Decisions take microseconds without an LLM call, at a domain-dependent benign-utility cost; compilation runs on-premise on open-weight gemma-4-26B.

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); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.11030 — 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 2026-10-09T05:47:01.490Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).