{"slug": "wire-profiling-witnessed-within-policy-instruction-collisions-in-llm-agents", "title": "WIRE: Profiling Witnessed Within-Policy Instruction Collisions in LLM Agents", "summary": "Researchers introduced WIRE, a profiler that detects collisions between standing rules within LLM agent prompt policies, finding that only 35.4% of 13,335 jointly governed trials satisfied both conflicting rules. Across six public prompt policies, WIRE extracted 276 source rules and 560 clauses, classified 30,944 clause-pair comparisons, and realized 1,402 concrete witnesses, revealing policy-specific, model-specific, and tool-interface-specific resolution patterns.", "body_md": "arXiv:2605.27784v2 Announce Type: replace\nAbstract: LLM agents are governed by long-lived prompt policies, where individually reasonable stand- ing rules can jointly govern the same pre- generation state. Existing instruction-following evaluations usually ask whether a model satis- fies explicit constraints, but they do not show how a model resolves pressure among rules inside one standing policy. We introduce WIRE, a witnessed resolu- tion profiler for prompt policies. WIRE ex- tracts source-grounded rules, encodes them as PYRULE clauses, uses satisfiability checks only to nominate same-surface hard-collision can- didates, realizes those candidates as concrete co-governance witnesses, and executes subject models to produce a four-cell resolution profile: satisfy both rules, only the earlier rule, only the later rule, or neither. Across six public prompt policies, WIRE ex- tracts 276 source rules and 560 clauses, clas- sifies 30,944 within-policy clause-pair com- parisons, retains 170 encoded hard-collision source-rule pairs, and realizes 1,402 concrete witnesses. In policy-only evaluation, these wit- nesses yield 13,335 jointly governed, judgeable trials; only 35.4% satisfy both governed rules. The resulting profiles reveal policy-specific, model-specific, and tool-interface-specific res- olution patterns. WIRE is not a proof of natural-language contra- diction, a deployment-frequency estimator, or a root-cause diagnosis. It is a measurement tool that returns reproducible witnesses and aggre- gate profiles for inspection, regression testing, and repair.", "url": "https://wpnews.pro/news/wire-profiling-witnessed-within-policy-instruction-collisions-in-llm-agents", "canonical_source": "https://www.machinebrief.com/news/wire-profiling-witnessed-within-policy-instruction-collision-te55", "published_at": "2026-07-28 04:00:00+00:00", "updated_at": "2026-07-28 05:57:28.124847+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-agents", "ai-safety"], "entities": ["WIRE", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/wire-profiling-witnessed-within-policy-instruction-collisions-in-llm-agents", "markdown": "https://wpnews.pro/news/wire-profiling-witnessed-within-policy-instruction-collisions-in-llm-agents.md", "text": "https://wpnews.pro/news/wire-profiling-witnessed-within-policy-instruction-collisions-in-llm-agents.txt", "jsonld": "https://wpnews.pro/news/wire-profiling-witnessed-within-policy-instruction-collisions-in-llm-agents.jsonld"}}