zenith6888/SkillsBench-1650
SkillBench-1650 A benchmark dataset for evaluating safety detection systems on AI agent skill packages. Contains 1,500 benign and 150 malicious Claude Code skill samples, designed for multi-dimensional security scoring research. Dataset Summary Split Samples Description benign 1,500 Real-world skills sampled from open-source repositories malicious 150 Adversarial payloads injected into real skill hosts Total 1,650 Benign samples are sourced… See the full description on the dataset page: https://huggingface.co/datasets/zenith6888/SkillsBench-1650.
| Author / organization | zenith6888 |
|---|---|
| License (as declared) | cc-by-4.0 |
| Downloads | 366 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 9 April 2026 |
| Last updated at source | 9 April 2026 |
Popularity, last 3 days
366 now (+4 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 17 Oct 2026 Description: SkillBench-1650 A benchmark dataset for evaluating safety detection systems on AI agent skill packages. Contains 1,500 benign and 150 malicious Claude Code skill samples, designed for multi-dimensional security scoring research. Dataset Sum → SkillBench-1650 A benchmark dataset for evaluating safety detection systems on AI agent skill packages. Contains 1,500 benign and 150 malicious Claude Code skill samples, designed for multi-dimensional security scoring research. Dataset Sum
Changes detected by Silytics between refreshes of the source.
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.