Security audit
Security checks for vulnerabilities and agentic risk
This is a simple Chinese performance and security assessment skill, but it asks for access/account details and describes penetration/load testing without enough safety scoping.
Use this only with systems you are authorized to test. Prefer staging targets and least-privilege test accounts, and do not paste privileged production credentials, API keys, session tokens, or customer data into the skill. Define scope, timing, rate limits, and rollback expectations before any load or penetration testing.
The description is written to prescribe output behavior in Chinese, including '输出量化评估结论和修复建议', but nowhere offers a language choice or states that Chinese is optional. Under the language/locale policy, a skill should not force a specific language unless the user opts in or the locale restriction is explicitly justified.
The skill explicitly asks for system access details, environment information, and accounts to perform security assessment, but it provides no safeguards on credential handling, no warning against sharing production secrets, and no recommendation to use scoped test accounts or non-production targets. In a security-testing skill, this omission increases the chance of unsafe secret exposure and accidental testing against live systems, which can lead to credential compromise or service disruption.
