SafeClawBench: Separating Semantic, Audit-Evidence, and Sandbox Harm in Tool-Using LLM Agents
Tool-using language-model agents introduce security failures that go beyond unsafe text: they can disclose protected objects, write persistent memory, send messages, modify databases, or trigger harmful code and tool effects. Existing evaluations often collapse these stages into a single attack success rate, making it difficult to tell whether a model merely agreed with an attacker or actually produced observable harm. We introduce SafeClawBench, a staged benchmark for tool-using agent security with 600 controlled adversarial tasks across six attack families: direct and indirect prompt injection, tool-return injection, memory poisoning, memory extraction, and ambiguity-driven unsafe infer...
| Authors | Yuchuan Tian, Mengyu Zheng, Haocheng Mei et al. |
|---|---|
| Primary category | cs.CR |
| Published | 16 June 2026 |
| Last updated at source | 16 June 2026 |
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.