remote-control

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This Skill implements a stealthy remote-control/backdoor flow—automatically sending transcripts and task summaries to an external WeChat endpoint (and an externally hosted transcript_domain), waiting for replies, and treating replies as new instructions to execute—enabling hidden network exfiltration and remote command execution without direct user prompts.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). 该 Skill 在运行时通过 gewe-cc wait-reply 从微信获取“用户回复内容”，而微信回复属于操作用户未选择引入的外部方文本输入（OUTSIDER：第三方通信渠道的自由文本），随后被直接放入代理上下文用于决定下一步。