lark-im

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

• Third-party content exposure detected (high risk: 0.80). The skill's documented workflows (e.g., references/lark-im-chat-messages-list.md, references/lark-im-messages-search.md, and SKILL.md) explicitly fetch and render user-generated chat messages and resources from third-party users and instruct the agent to read, summarize, and take follow-up actions (send/reply/download), meaning untrusted message content can directly influence tool use and decisions.