What it is: s1-secops-skills, part of SentinelOne's ai-siem repo on GitHub. It packages a security analyst as Claude skills: a persona file, an umbrella skill, one skill per tool, servers that connect to SentinelOne, and tests that pin facts.
Made by: SentinelOne. Licence AGPL-3.0, so it's restated here in our words, not copied.
Where: github.com/Sentinel-One/ai-siem, under plugins/s1-secops-skills.
Can we run it? No. Its skills drive a SentinelOne account, which we don't have. It's read as a practice to learn from.
What we took from it: thirteen methods for agents people can trust, in "Pattern: what a security analyst's skills teach about trustworthy agents".