alibabacloud-aes-sysom-lingjun-diagnosis

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill instructs the agent to request a DingTalk webhook URL from the user and then embed that user-provided webhook URL verbatim into a command/SDK-call (scripts/create-alert-destination.py ''), which requires the LLM to handle and output a secret-like token directly.

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

• Third-party content exposure detected (medium risk: 0.30). In the alert configuration Phase 3, the runtime reads user-supplied free text “Webhook URL” (Step 10) and passes it into the SDK script scripts/create-alert-destination.py, which then creates a DingTalk alert destination using that untrusted string.