SafeClawArena Benchmark

What is SafeClawArena?

SafeClawArena evaluates security failures in Claw-like agents using 406 adversarial tasks spanning four attack surfaces in containerized platform replicas.

Released

2026-06-29

Evaluates

Agents, Cybersecurity, Language & Knowledge

Openness

unknown

Importer

Claire Radar

Review status

ai-reviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2606.30755

SafeClawArena paper and code

• Paper https://arxiv.org/abs/2606.30755
• Code https://github.com/sunblaze-ucb/SafeClawArena

Which sources cite SafeClawArena?

• SafeClawArena

Related Agents, Cybersecurity, Language & Knowledge benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• A3S-Bench
• AARRI-Bench
• ABC-Bench
• ActiveSciBench
• AdaMem-Bench