Security audit
Security checks for vulnerabilities and agentic risk
This is a finance-analysis Markdown skill with broad wording and tool permissions, but I found no hidden execution, persistence, exfiltration, or destructive behavior.
Install only if you want an agent to help with finance-analysis workflows. Because it declares read, exec, and write, review any generated commands or report-writing actions before allowing them, and provide financial data or API credentials only when needed.
The description includes a very broad activation cue ('Use when 用户需要财务分析入门工具相关功能时使用') that can match almost any finance-related request without clear boundaries. In an agent environment with read/exec/write tools, this increases the chance the skill is invoked for vague prompts and gains access to file or command capabilities unnecessarily.
The capability-scope section lists expansive natural-language triggers such as data analysis, report generation, statistical insights, and data visualization, which go well beyond the declared finance focus. This broad phrasing can cause over-selection of the skill for generic analytical requests, making the presence of exec/write tools more dangerous because the agent may choose this skill in contexts where those capabilities are unnecessary.
