The paper introduces Many-Tier Instruction Hierarchy (ManyIH), a new framework for resolving conflicts among instructions with arbitrarily many privilege levels in large language model agents. It presents ManyIH-Bench, a benchmark featuring 853 agentic tasks that require navigating up to 12 levels of conflicting instructions across 46 real-world agents. Experiments show current models achieve only about 40% accuracy when instruction conflict scales, highlighting a gap in fine-grained, scalable conflict resolution.
arXiv:2511. 04694v5 Announce Type: replace-cross Abstract: As large language model (LLM) based systems take on high-stakes roles in real-world decision-making, they must reconcile competing instructions from multiple sources within a single prompt context.
arXiv:2606. 07808v1 Announce Type: new Abstract: Reasoning language models deployed in agentic workflows must follow an instruction hierarchy: when instructions from different sources conflict, the model should obey the highest-privilege applicable instruction.
SkillEffect is a checked‑lowering runtime that ensures agent tool calls stay within memory limits by verifying each proposed program against an immutable input before execution. It uses audited relation plugins to provide source recognition, bounded intermediate representation construction, and postconditions, while a shared runtime handles selection, bounded VM execution, and atomic capacity leasing. Experiments across six operator families show that bounded access significantly reduces peak memory usage and improves completion rates under fixed memory caps.
Agent Skills can specify procedural and resource obligations for tool use, and language models instantiate them as concrete programs. However, when models turn this guidance into code for existing too...
arXiv:2602. 12430v4 Announce Type: replace-cross Abstract: The transition from monolithic language models to modular, skill-equipped agents marks a defining shift in how large language models (LLMs) are deployed in practice.
arXiv:2603. 17673v2 Announce Type: replace-cross Abstract: LLM agents are becoming increasingly important in the security domain, but leading systems are often closed-source, cloud-based, hard to reproduce or use with sensitive code.
The paper introduces AgentLeak, a black‑box attack that clones the task‑solving capabilities of a strong LLM agent onto a weaker one by exploiting differences between successful and failed executions. Unlike prior skill‑stealing methods that only recover explicit skill artifacts, AgentLeak identifies and incorporates missing procedural behaviors, boosting task pass rates by over 40% and closing more than 80% of the capability gap across 20 scenarios. The study demonstrates that observable execution behavior can leak proprietary procedural knowledge, posing a confidentiality risk for LLM agents.
arXiv:2609.14992v1 Announce Type: new Abstract: Recently, the rapid development of large language models (LLMs) has reshaped software engineering by enabling autonomous code agents that plan, execute...
arXiv:2605. 09163v3 Announce Type: replace Abstract: Large language model agents increasingly operate through an intermediate skill layer that mediates between user intent and concrete task execution.
arXiv:2608.30041v1 Announce Type: cross Abstract: Large language model agents place outputs from external skills into their execution context, allowing attacker-controlled data to influence later pri...
The paper introduces a post‑training framework that teaches a 4B‑parameter language model to exercise task‑conditioned authority in executable terminal and Model Context Protocol (MCP) environments. By auditing each action across six risk dimensions with deterministic verifiers and optimizing for task‑specific excess‑privilege values, the authors achieve 98.48% safe success and reduce excess‑authority errors from 4.56% to 0.79% on held‑out tasks. The study also demonstrates capability retention, prompt‑directed improvement, and generalization over a 400‑task continuation test.
ClawSentry is an open‑source, framework‑agnostic security supervision gateway designed to protect autonomous large language model (LLM) agents from progressive risks that can arise at four points in the agent control loop: skill admission, invocation‑time intent, execution‑time effect, and post‑action consequence. It introduces a multi‑tier decision engine—deterministic L1, rule‑anchored L2, and read‑only L3—alongside a First‑Use Skill Package Review (FSPR) and an Agent Harness Protocol (AHP) that applies a single policy across multiple agent runtimes without modifying their internals. Evaluation on SkillInject and the SkillsSafety benchmark shows that ClawSentry significantly reduces contextual adversarial skill risk (ASR) while maintaining high task success rates (TSR).