Newest papers

Certified Execution Harness for Agentic AI: Safe and Least-Intrusive Petri-Net Supervisory Control Between Reasoning and Action

Chao Gu, Quanyan Zhu · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.31376 · cited by 0 works (OpenAlex) · open-access PDF

Hindi

इस दस्तावेज़ के चरण में, हम त्वरित इंजेक्शन के तहत बड़े भाषा मॉडल (एलएलएम) एजेंटों के लिए सुरक्षित उपकरण निष्पादन की समस्या का अध्ययन करते हैं। Petri नेट का उपयोग करके, हम एलएलएम और इसके उपकरणों के बीच एक निगरानी नियंत्रण समस्या के रूप में निष्पादन हार्डवेयर को औपचारिक बनाते हैं, जिसमें मार्केटिंग क्षमताओं, अनुमोदनों और बजटों का प्रतिनिधित्व करते हैं। चूंकि एक संक्रमण केवल तभी आग लगती है जब इसकी आवश्यकता होती है, एक इंजेक्शन अनुरोध जिसकी पूर्वानुमानों को पूरा नहीं किया जाता है या पहले से ही खर्च किया जाता है, जिसका अर्थ के बावजूद इसे अस्वीकार किया जाता है। उपकरण के परिणामों को नियंत्रित करने के लिए, हम पारंपरिक अधिकतम अनुमतिदार

Machine translation (Meta M2M100); the English original follows.

Abstract

This paper studies the problem of safe tool execution for large language model (LLM) agents under prompt injection. Using Petri nets, we formalize the execution harness between the LLM and its tools as a supervisory control problem, in which marks represent capabilities, approvals and budgets. Since a transition fires only if the marks it needs are present, an injected request whose preconditions are unmet or already spent is refused regardless of its wording. Treating tool outcomes as uncontrollable, we enforce safety by the conventional maximally permissive supervisor, and this safety transfers to the running system under four assumptions. Moreover, to capture an agent that may propose again after a refusal, we introduce the notion of a replanning closed loop, in which each retry costs an inference call. We show that, in this loop, the conventional supervisor can be arbitrarily more intrusive than a stricter safe one. Furthermore, for a given proposer model, the maximally permissive least-intrusive supervisor is characterized as a value-threshold refinement of the conventional one. Under precise conditions, this refinement orders two concurrent actions by at most one monitor place. On an autonomous release agent, we find by synthesis that safe deployment requires a step that the supervisor can refuse after a rollout starts.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Supervisor, Supervisory control, Action (physics), Computer science, Petri net, Control (management)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange