ClawsBench - Unsafe Action Rate (Full Scaffold): leaderboard
Unsafe actions on the 24 safety-critical tasks of ClawsBench (2026), where an LLM agent works in stateful mock Gmail, Slack, Calendar, Docs and Drive services through the OpenClaw harness with domain skills and a meta prompt enabled, five repeats per task: the share of trials in which the agent takes an irreversible harmful action, such as leaking confidential data or deleting protected records. Percent; lower is better.
Metric: Unsafe action rate (%): share of trials on the 24 safety-critical tasks with a negative score (an irreversible harmful action such as leaking confidential data or deleting protected records), ClawsBench's 44 tasks over five stateful mock services (Gmail, Slack, Calendar, Docs, Drive), OpenClaw harness with domain skills and the meta prompt both enabled (full scaffolding), 5 repeats per task, state-based programmatic scoring; lower is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 7 | #135 |
| 2 | Gemini 3.1 Pro (Preview) | 10 | #99 |
| 3 | Claude Sonnet 4.6 | 13 | #173 |
| 4 | Claude Opus 4.6 | 23 | #113 |
| 5 | GLM-5 | 23 | #269 |
| 6 | Gemini 3.1 Flash Lite | 23 | #371 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=clawsbench-unsafe-action-rate-full-scaffold · How It Works · Data refreshed daily, snapshot 2026-11-21.