implement-tmux

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

• Third-party content exposure detected (medium risk: 0.30). 运行时通过 tmux send-keys 将由外部提交生成的 map.md/spec.md（来自 /wayfinder//to-tickets 的输入）内容嵌入到 worker 的首次 prompt，并由 worker 在其会话中执行 /implement；因此外部可注入的自由文本会进入 LLM（worker agent）上下文。

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly instructs creating and killing tmux windows, running an agent CLI with the --permission-mode bypassPermissions flag (bypassing security), and driving commits/verification in the project—directly modifying the machine/session state and requesting a permission-bypass, which is a high-risk behavior.