Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed handbook-building workflow that reads user-provided materials and writes organized Markdown manuals without hidden execution or persistence.
Install this if you want an assistant to turn selected Markdown files and links into a maintained handbook. Before using it, point it only at directories and URLs you intend the assistant to read, because it may process every .md file in the specified directory and incorporate source material into generated handbook files.
The trigger description is overly broad, including generic requests like '帮我整理这些文档' or '总结这些材料', which can cause the skill to activate in contexts where the user did not intend multi-file handbook generation. This increases the chance of unnecessary file access and structural transformations, especially because the skill also instructs reading many local files and external URLs once activated.
The skill explicitly instructs reading all .md files in a user-specified directory and fetching content from supplied URLs, but it does not require user-facing notice or confirmation about local data access and outbound network retrieval. This can lead to unintended exposure of sensitive local documentation, retrieval of untrusted remote content, or privacy/security issues if the user does not realize the scope of access.
No suspicious patterns detected.
