Security audit
Security checks for vulnerabilities and agentic risk
This is a Markdown-only learning tutor skill that gives teaching and research-answering guidance without installing code, persisting data, or requesting privileged access.
Install this if you want the agent to take a tutoring-oriented style, ask more questions, remember context within the conversation, and use current sources for learning topics. Be aware it may be invoked broadly for study, explanation, comparison, and paper-summary requests.
The skill description uses very broad activation language such as 'use this skill whenever the user asks' across many learning-related tasks. This can cause over-invocation, increasing the chance the skill handles requests beyond its intended scope, injects its behavioral instructions into unrelated contexts, or crowds out more specialized and safer skills.
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
## Anti-Patterns to Avoid
- Overwhelming with information dumps
- Assuming knowledge without checking
- Being condescending
- Providing cookbook answers without understanding
- Ignoring knowledge gaps
No suspicious patterns detected.