Security audit
Security checks for vulnerabilities and agentic risk
The skill is mostly coherent for a daily social-media workflow, but it can access logged-in social accounts and may change account state without a clear confirmation step.
Install only if you are comfortable with the agent viewing logged-in social-media pages, notifications, profile metrics, and engagement context. Before running it, require scan-only behavior by default and confirm separately before accepting LinkedIn connections, posting comments, replying, or writing tracking data.
The manifest description explicitly says the routine runs across four platforms including 小红书, and metadata also lists xiaohongshu as a platform. However, the execution flow, action items, browser automation, database tracking, and weekly review only define concrete steps for Reddit, Twitter/X, and LinkedIn, with no operational instructions for 小红书.
The trigger phrases include broad natural-language expressions like "social media check" and "do my social today," which can match ordinary conversation and invoke browser automation unexpectedly. In a skill that checks notifications, engagement data, and account activity across multiple platforms, accidental activation can lead to unintended access to authenticated sessions and actions based on private account context.
The skill instructs use of Playwright/browser tools to visit Reddit and Twitter/X account pages and notifications, but it does not clearly warn users that it may access logged-in sessions, platform data, and potentially sensitive notifications. Without explicit disclosure and consent, users may unknowingly grant the skill access to personal or business social-media content, increasing the risk of privacy violations and unintended account interactions.
