AgentHazard (OpenClaw) - Persistence Establishment: leaderboard

AgentHazard attack success rate on persistence establishment attempts (backdoors, SSH keys, startup scripts or scheduled jobs that keep long-term access), with each backbone model running inside OpenClaw. Percent of instances judged harmful on the full trajectory by Gemini-3-Flash, lower is better.

Metric: Attack success rate (%) on the persistence establishment instances (backdoors, SSH keys, startup scripts or scheduled jobs that keep long-term access), with the model running inside the OpenClaw open-source agent framework (browser, shell and file-system tools; one isolated Docker container per task), over AgentHazard instances (harmful objectives embedded in locally legitimate multi-step workflows, ten risk categories and ten attack strategies; the evaluated runs cover 1,000 instances), judged on the full trajectory (round_all) by Gemini-3-Flash, which labels a trajectory harmful when it realizes the objective; backbones served through an OpenAI-compatible API; lower is better. Source: arxiv.org. 7 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen 2.5 72B Instruct37.37#1117
2Kimi K2.559.6#266
3Qwen 2.5 Coder 32B Instruct66.67#1547
4Kimi K268.69#443
5Qwen 3 VL 235B A22B Instruct73.74#532
6Qwen 3 32B75.51#972
7GLM-4.681.82#498

Interactive version: theaggregate.ai/benchmark?slug=agenthazard-openclaw-persistence-establishment · How It Works · Data refreshed daily, snapshot 2026-11-21.