A cross-layer runtime analysis framework, Oscar, attributes known OSS vulnerability effects in concrete agent executions and localizes the furthest layer (runtime, tool, or model-visible observation) at which the effect remains manifest.
AGENTVIGIL : Automatic Black-Box Red-teaming for Indirect Prompt Injection against LLM Agents
4 Pith papers cite this work, alongside 2 external citations. Polarity classification is still indexing.
4
Pith papers citing it
2
external citations · external index
citation-role summary
background 3
citation-polarity summary
years
2026 4roles
background 3representative citing papers
A co-evolutionary loop pairing latency-guided attack mining with reward-decoupled policy optimization cuts indirect prompt injection success by 88.5 percent on three benchmarks.
citing papers explorer
-
Characterizing Security Effects of OSS Vulnerabilities in Agent Systems
A cross-layer runtime analysis framework, Oscar, attributes known OSS vulnerability effects in concrete agent executions and localizes the furthest layer (runtime, tool, or model-visible observation) at which the effect remains manifest.
-
CoDeL: Co-Evolutionary Defense against Indirect Prompt Injection in LLM-based Agents
A co-evolutionary loop pairing latency-guided attack mining with reward-decoupled policy optimization cuts indirect prompt injection success by 88.5 percent on three benchmarks.
- Toward Secure LLM Agents: Threat Surfaces, Attacks, Defenses, and Evaluation
- SkillHarm: Lifecycle-Aware Skill-Based Attacks via Automated Construction