Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed prompt-optimization helper with no code execution, external access, persistence, or hidden data handling.
This appears safe to install for prompt optimization. Be aware it may activate on broad prompt-refinement requests, and you should review the optimized prompt before using it if wording, language, or constraints matter.
The trigger description is very broad and matches generic requests such as optimization, refinement, and token saving across many domains. This can cause the skill to activate when the user did not explicitly request this skill, leading to prompt hijacking of normal conversations and unintended instruction shaping.
The skill claims applicability across a very wide range of scenarios without defining activation boundaries or disambiguation rules. In an agent environment, this increases the chance of unintended invocation over unrelated tasks, which can override more appropriate skills and alter user requests in ways the user did not ask for.
The text says the skill will automatically detect Chinese/English and maintain language consistency, which imposes a language behavior automatically. Because it does not offer the user a language preference or opt-in, it may violate language or locale choice expectations.
The instruction to output Chinese if input is Chinese and English if input is English is a fixed language policy. It does not mention any user override, so it constrains language choice without explicit consent.
