Spider-Sense: Intrinsic Risk Sensing for Efficient Agent Defense with Hierarchical Adaptive Screening

TL;DR

Spider-Sense introduces an intrinsic, event-driven security framework for autonomous agents that selectively triggers defenses based on risk perception, improving efficiency and effectiveness over traditional mandatory checks.

Contribution

It proposes a novel Intrinsic Risk Sensing framework with hierarchical defense, and introduces S$^2$Bench, a benchmark for evaluating agent security against multi-stage attacks.

Findings

Achieves lowest Attack Success Rate (ASR) and False Positive Rate (FPR)

Maintains only 8.3% latency overhead

Outperforms existing defense mechanisms in experiments

Abstract

As large language models (LLMs) evolve into autonomous agents, their real-world applicability has expanded significantly, accompanied by new security challenges. Most existing agent defense mechanisms adopt a mandatory checking paradigm, in which security validation is forcibly triggered at predefined stages of the agent lifecycle. In this work, we argue that effective agent security should be intrinsic and selective rather than architecturally decoupled and mandatory. We propose Spider-Sense framework, an event-driven defense framework based on Intrinsic Risk Sensing (IRS), which allows agents to maintain latent vigilance and trigger defenses only upon risk perception. Once triggered, the Spider-Sense invokes a hierarchical defence mechanism that trades off efficiency and precision: it resolves known patterns via lightweight similarity matching while escalating ambiguous cases to deep internal reasoning, thereby eliminating reliance on external models. To facilitate rigorous evaluation, we introduce S$^2$Bench, a lifecycle-aware benchmark featuring realistic tool execution and multi-stage attacks. Extensive experiments demonstrate that Spider-Sense achieves competitive or superior defense performance, attaining the lowest Attack Success Rate (ASR) and False Positive Rate (FPR), with only a marginal latency overhead of 8.3\%.