Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese-language skill for reviewing and improving one skill file, with no bundled code, install hook, or hidden execution behavior found.
Use this only when you intentionally want an agent to evaluate and modify a skill instruction file. Provide the exact target SKILL.md, require review of the diff and test results, and confirm before retaining changes, especially because modifying skills can affect future agent behavior.
The file presents the skill entirely in Chinese and does not indicate that the user may choose another language or that the locale restriction is required for a region-specific purpose. Under the policy, forcing a specific language without opt-in is a natural-language policy concern.
The invocation guidance is overly broad ('optimize/evaluate existing skills') and does not define clear scope, allowed targets, or stopping conditions. In an autonomous optimization skill, vague triggers can cause the agent to act on unintended skills or repositories, increasing the risk of unauthorized modification, prompt-scope drift, or recursive self-modification without sufficient user intent verification.
No suspicious patterns detected.
