PrefixGuard: From LLM-Agent Traces to Online Failure-Warning Monitors

• URL: http://arxiv.org/abs/2605.06455v1
• Date: Thu, 07 May 2026 15:49:48 GMT
• Title: PrefixGuard: From LLM-Agent Traces to Online Failure-Warning Monitors
• Abstract summary: Large language model (LLM) agents execute long, tool-using tasks where final outcome checks can arrive too late for intervention.<n>We introduce PrefixGuard, a trace-to-monitor framework with an offline StepView induction step followed by supervised monitor training.<n>Across WebArena, $2$-Bench, SkillsBench, and TerminalBench, the strongest PrefixGuard monitors reach 0.900/0.70.533/0.557 AUPRC.
• Score: 14.336100401626062
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language model (LLM) agents now execute long, tool-using tasks where final outcome checks can arrive too late for intervention. Online warning requires lightweight prefix monitors over heterogeneous traces, but hand-authored event schemas are brittle and deployment-time LLM judging is costly. We introduce PrefixGuard, a trace-to-monitor framework with an offline StepView induction step followed by supervised monitor training. StepView induces deterministic typed-step adapters from raw trace samples, and the monitor learns an event abstraction and prefix-risk scorer from terminal outcomes. Across WebArena, $τ^2$-Bench, SkillsBench, and TerminalBench, the strongest PrefixGuard monitors reach 0.900/0.710/0.533/0.557 AUPRC. Using the strongest backend within each representation, they improve over raw-text controls by an average of +0.137 AUPRC. LLM judges remain substantially weaker under the same prefix-warning protocol. We also derive an observability ceiling on score-based area under the precision-recall curve (AUPRC) that separates monitor error from failures lacking evidence in the observed prefix. For finite-state audit, post-hoc deterministic finite automaton (DFA) extraction remains compact on WebArena and $τ^2$-Bench (29 and 20 states) but expands to 151 and 187 states on SkillsBench and TerminalBench. Finally, first-alert diagnostics show that strong ranking does not imply deployment utility: WebArena ranks well yet fails to support low-false-alarm alerts, whereas $τ^2$-Bench and TerminalBench retain more actionable early alerts. Together, these results position PrefixGuard as a practical monitor-synthesis recipe with explicit diagnostics for when prefix warnings translate into actionable interventions.

Related papers

• Automata from Agent Traces: Failure and Next-Step Prediction [1.16693435589939]
  We collapse an entire trace corpus into a single, compact finite-state machine (FSM) that serves as a structural substrate for the behavior of LLM agents.<n>For failure prediction, per-state behavioral features reach held-out AUROC up to 0.94, and an online monitor ranks failing runs above passing ones.
  arXiv  Detail & Related papers  (2026-08-24T17:58:01Z)
• From Attack Simulation to SIEM Rule: Deterministic Detection-as-Code Synthesis with Probe-Level Traceability [51.56484100374058]
  Security teams simulate attacks against their own systems to check whether their monitoring would catch a real intruder.<n>A human bridges that gap by hand, reading each finding and writing the corresponding Sigma rule.<n>We show this translation can be partially automated when probes are drawn from a locked corpus.
  arXiv  Detail & Related papers  (2026-06-03T14:26:25Z)
• Early Diagnosis of Wasted Computation in Multi-Agent LLM Systems via Failure-Aware Observability [8.036549927091286]
  This paper introduces a failure-aware observability framework for diagnosing wasted computation in multi-agent LLM traces.<n>We instantiate the framework in a three- agent question-answering system and evaluate it on 165 GAIA validation traces under identical execution caps.
  arXiv  Detail & Related papers  (2026-05-31T17:50:11Z)
• $D^2$-Monitor: Dynamic Safety Monitoring for Diffusion LLMs via Hesitation-Aware Routing [63.49501120848927]
  Safety monitoring for large language models (D-LLMs) remains largely unexplored.<n>We propose $D2$-Monitor, a bi-level safety monitor for D-LLMs.<n>$D2$-Monitor adopts a lightweight probe as an always-on monitor to jointly estimate hesitation and perform base classification.
  arXiv  Detail & Related papers  (2026-05-25T14:22:21Z)
• SLEIGHT-Bench: A Benchmark of Evasion Attacks Against Agent Monitors [3.989565392866092]
  We present SLEIGHT-Bench (Subtle Low-itEration Insight-Guided Harmful Transcripts), a benchmark of synthetic transcripts containing 40 attacks across 11 categories.<n>We find that 20 of 40 attacks are never caught by an Opus 4.6 monitor with extended thinking across 10 trials at a 1% false-positive rate, while the overall catch rate is 32%.
  arXiv  Detail & Related papers  (2026-05-15T20:46:46Z)
• ShieldNet: Network-Level Guardrails against Emerging Supply-Chain Injections in Agentic Systems [56.613157564882925]
  Malicious behaviors are embedded in seemingly benign tools, silently hijacking agent execution, leaking sensitive data, or triggering unauthorized actions.<n>Despite their growing impact, there is currently no comprehensive benchmark for evaluating such threats.<n>We propose ShieldNet, a network-level guardrail framework that detects supply-chain poisoning by observing real network interactions.
  arXiv  Detail & Related papers  (2026-04-06T05:15:00Z)
• DeceptGuard :A Constitutional Oversight Framework For Detecting Deception in LLM Agents [0.0]
  We introduce DECEPTGUARD, a unified framework that systematically compares three monitoring regimes.<n>We introduce DECEPTSYNTH, a scalable synthetic pipeline for generating deception-positive and deception-negative agent trajectories.<n>Our monitors are optimized on 4,800 synthetic trajectories and evaluated on 9,200 held-out samples from DeceptArena.
  arXiv  Detail & Related papers  (2026-03-14T06:45:43Z)
• Detecting Object Tracking Failure via Sequential Hypothesis Testing [80.7891291021747]
  Real-time online object tracking in videos constitutes a core task in computer vision.<n>We propose interpreting object tracking as a sequential hypothesis test, wherein evidence for or against tracking failures is gradually accumulated over time.<n>We propose both supervised and unsupervised variants by leveraging either ground-truth or solely internal tracking information.
  arXiv  Detail & Related papers  (2026-02-13T14:57:15Z)
• Adaptive Attacks on Trusted Monitors Subvert AI Control Protocols [80.68060125494645]
  We study adaptive attacks by an untrusted model that knows the protocol and the monitor model.<n>We instantiate a simple adaptive attack vector by which the attacker embeds publicly known or zero-shot prompt injections in the model outputs.
  arXiv  Detail & Related papers  (2025-10-10T15:12:44Z)
• Beyond Linear Probes: Dynamic Safety Monitoring for Language Models [67.15793594651609]
  Traditional safety monitors require the same amount of compute for every query.<n>We introduce Truncated Polynomials (TPCs), a natural extension of linear probes for dynamic activation monitoring.<n>Our key insight is that TPCs can be trained and evaluated progressively, term-by-term.
  arXiv  Detail & Related papers  (2025-09-30T13:32:59Z)
• BlindGuard: Safeguarding LLM-based Multi-Agent Systems under Unknown Attacks [58.959622170433725]
  BlindGuard is an unsupervised defense method that learns without requiring any attack-specific labels or prior knowledge of malicious behaviors.<n>We show that BlindGuard effectively detects diverse attack types (i.e., prompt injection, memory poisoning, and tool attack) across multi-agent systems.
  arXiv  Detail & Related papers  (2025-08-11T16:04:47Z)
• CoT Red-Handed: Stress Testing Chain-of-Thought Monitoring [3.6284577335311563]
  Chain-of-Thought (CoT) monitoring improves detection by up to 27 percentage points in scenarios where action-only monitoring fails to reliably identify sabotage.<n>CoT traces can also contain misleading rationalizations that deceive the monitor, reducing performance in more obvious sabotage cases.<n>This hybrid monitor consistently outperforms both CoT and action-only monitors across all tested models and tasks, with detection rates over four times higher than action-only monitoring for subtle deception scenarios.
  arXiv  Detail & Related papers  (2025-05-29T15:47:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.