engineering-feishu-integration-developer

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow ingests Feishu event payloads (e.g., im.message.receive_v1 and approval.approval.updated_v4) delivered to the webhook handlers, where message.content / callback data originates from other users/parties in Feishu and is parsed into LLM context via the agent’s handling logic (outsider-authored free text/comments).