SkillSonar generalizes zero-shot to external safety benchmarks, achieving the lowest ASR of 52.90% on SkillSafetyBench and highest safety score of 54.00% on WildClawBench compared to prompt, permission, and allowlist baselines.

SkillSonar generalizes zero-shot to external safety benchmarks, achieving the lowest ASR of 52.90% on SkillSafetyBench and highest safety score of 54.00% on WildClawBench compared to prompt, permission, and allowlist baselines. · CiteArk

Measurement results

0/9 assessed

0 supported0 disputed or conflicting9 not assessed

The complete result group stays with its claim. Assessed rows come first; their judgments do not replace the whole-claim conclusion. Expand a row for conditions, sources, and evidence history.

Each row is a measurement with its reported value, latest available observation, and cumulative evidence state.
Metric and conditions	Reported	Latest observation	Evidence judgment
defense: No-guard · benchmark: SkillSafetyBench	93.3 percentage_points	—	Not assessed
defense: No-guard · benchmark: WildClawBench	38 percentage_points	—	Not assessed
defense: AcceptEdits · benchmark: WildClawBench	41 percentage_points	—	Not assessed
defense: AcceptEdits + allowlist · benchmark: SkillSafetyBench	64.3 percentage_points	—	Not assessed
defense: AcceptEdits + allowlist · benchmark: WildClawBench	45 percentage_points	—	Not assessed
defense: System prompt · benchmark: SkillSafetyBench	83.3 percentage_points	—	Not assessed
defense: System prompt · benchmark: WildClawBench	43 percentage_points	—	Not assessed
defense: SkillSonar (ours) · benchmark: SkillSafetyBench	52.9 percentage_points	—	Not assessed
defense: SkillSonar (ours) · benchmark: WildClawBench	54 percentage_points	—	Not assessed

SkillSonar generalizes zero-shot to external safety benchmarks, achieving the lowest ASR of 52.90% on SkillSafetyBench and highest safety score of 54.00% on WildClawBench compared to prompt, permission, and allowlist baselines.

Reported and observed measurements

Assessments (0)

Runs (0)

Measurement results