Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese CCA study skill that teaches agent architecture and may create practice code, with no evidence of hidden, destructive, persistent, or data-exfiltrating behavior.
Install this if you want Chinese CCA domain 1 tutoring. Before using the hands-on exercise in an important repository, ask the agent to show the planned files first or work in a dedicated example directory so sample code does not mix with production code.
The instructional content is written as mandatory Chinese-language tutoring and quiz behavior, with no indication that the user may choose another language. This can violate language or locale policy when a skill imposes a specific language without user opt-in.
The skill explicitly tells the agent to help create example code in the current project, yet it provides no guardrails about asking permission before modifying files, using a safe workspace, or avoiding overwriting user code. In an environment with Write/Edit/Bash privileges, this can lead to unintended file modification, project corruption, or insertion of code the user did not clearly authorize.
The description says the skill should be used when the user says phrases like "学domain1", "代理架构", or "agent编排". While some are domain-related, the activation guidance does not define boundaries or negative examples, so phrases like "代理架构" could match general discussion rather than an explicit request to invoke this skill.
No suspicious patterns detected.
