Security audit
Security checks for vulnerabilities and agentic risk
This skill is a simple memory-classification guide with disclosed storage recommendations and no executable code, though users should be mindful about what they choose to persist.
Install only if you want help structuring notes for possible memory storage. Before accepting any recommended storage action, avoid saving secrets or sensitive personal data unless you explicitly intend to, and review whether items should be stored, updated, held, or kept ephemeral.
The activation triggers are broad and subjective, so an agent could invoke this skill in many contexts involving notes, memory, or retrieval without clear gating. Because the skill explicitly touches memory handling and user context, over-invocation can cause unnecessary persistence, misclassification of sensitive information, or use outside the user's intent.
The skill references storing user context safely for future retrieval but provides no privacy warning, consent requirement, retention guidance, or limits on what may be stored. In a memory-oriented skill, this omission is meaningful because users or downstream agents may infer that personal, sensitive, or provisional information can be persisted by default.
The default prompt is broad enough to match many ordinary 'classify this information' requests, and with implicit invocation enabled this can cause the skill to activate without clear user intent. That increases the chance of unintended memory processing or storage behavior, especially when users are asking for generic summarization or organization rather than durable memory creation.
