Security audit
Security checks for vulnerabilities and agentic risk
This is a simple Xiaohongshu-style copywriting skill with no code execution, data access, persistence, or hidden behavior.
Install this if you want stylized Xiaohongshu promotional copy. Review generated content for audience fit, exaggeration, and platform policy compliance before publishing, since the skill is designed to produce attention-grabbing marketing text.
The trigger description is broad enough to match ordinary writing requests, which can cause the skill to activate outside its intended niche. Over-broad activation can override user intent, route benign requests through a manipulative/social-media-optimized persona, and create unreliable or unsafe automation behavior in agent pipelines.
The skill hardcodes a specific 'bestie-like' tone and female-targeted persuasive style without allowing user choice or context checks. This is risky because it can produce manipulative, biased, or misaligned output when the user's audience, platform, or preferences differ, reducing user control and potentially encouraging deceptive engagement tactics.
