Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward test-case generator that reads user-provided requirements and writes structured markdown test documentation.
Before installing, understand that the skill may read PRD files you point it to and save generated markdown under tests/. Ask the agent to confirm the target path, avoid overwriting existing files, and summarize in your preferred language if Chinese is not suitable.
The skill explicitly directs the agent to write output to tests/<name>-test-cases.md without requiring user confirmation or clearly surfacing that a filesystem modification will occur. In an agentic environment, silent file creation or overwrite can cause unintended workspace changes, clobber existing artifacts, or be abused to persist misleading content without the user's informed consent.
Mandating that the final summary be in Chinese removes user choice and can reduce transparency if the user does not read Chinese. In security-sensitive or workflow-sensitive contexts, this can obscure what was generated, what assumptions were made, and where files were written, increasing the chance that users approve or miss side effects they do not fully understand.
No suspicious patterns detected.
