lark-minutes

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

• Third-party content exposure detected (medium risk: 0.30). 该 skill 的运行路径通过 minutes +search 等接口由用户所指定的搜索文本/过滤条件发起妙记检索（POST /open-apis/minutes/v1/minutes/search），会在运行时摄取外部输入的自由文本（如 --query）用于搜索与后续选择命中条目，从而存在间接提示注入风险。

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). SKILL.md explicitly requires using a Read tool at runtime to fetch ../lark-shared/SKILL.md and ../lark-vc/references/vc-domain-boundaries.md (SKILL.md lines 3 and 5), and those files are mandatory inputs that control authentication/permission and domain-boundary behavior for the agent, so they directly influence agent instructions.