ClawsBench - Safety (Gemini CLI; skills off; meta on): leaderboard

Metric: Mean native task score x100 (-100 to 100); Gemini CLI, domain skills off, meta prompt on; mean over observed scored trials, including scored error rows; higher is better. Source: huggingface.co. 1 models tracked.

Top models

#ModelScore
1gemini-3.1-flash-lite-preview2.68

Interactive version: theaggregate.ai/benchmark?slug=clawsbench-safety-gemini-cli-skills-off-meta-on · How It Works · Data refreshed daily, snapshot 2026-10-09.