Security audit
Security checks for vulnerabilities and agentic risk
This is a markdown-only geoscience reasoning skill that may trigger broadly but does not request sensitive access, execute code, or persist changes.
Installers should expect this skill to influence answers about geology and subsurface topics, and occasionally unrelated uses of words like pressure, depth, formation, or structure. It does not appear to run code or access data, but its routing triggers could be narrowed for cleaner behavior.
The activation criteria are very broad, including generic conditions like 'any claim involves' broad Earth-domain topics and a mandatory-sounding activation. In a multi-skill agent, this can cause inappropriate triggering on loosely related prompts, leading to skill overreach, degraded routing accuracy, and potentially incorrect constraint application to non-geoscience tasks.
The keyword trigger list contains ambiguous terms such as 'formation', 'structure', 'pressure', and 'depth' that are common across many domains. If used as activation signals, they can spuriously invoke the skill in unrelated contexts, causing misrouting and contaminating responses with irrelevant domain rules.
