alibabacloud-dsc-audit

Pass

Audited by Gen Agent Trust Hub on Apr 28, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill includes defensive instructions that specifically prohibit the agent from echoing, printing, or requesting sensitive Alibaba Cloud Access Keys (AK/SK), which helps protect against credential harvesting through prompt manipulation.
  • [COMMAND_EXECUTION]: The skill utilizes Python scripts to perform administrative tasks. The handle_risk.py script includes comprehensive input validation using regular expressions and length checks to prevent command injection and ensure that user-supplied handling details are sanitized before being sent to the API.
  • [EXTERNAL_DOWNLOADS]: The skill requires official Alibaba Cloud Python SDK packages (e.g., alibabacloud_tea_openapi) from the standard PyPI registry. These are legitimate dependencies provided by the vendor.
  • [SAFE]: All network operations are directed exclusively to official Alibaba Cloud endpoints (sddp.cn-zhangjiakou.aliyuncs.com), and credential management relies on the official CredentialClient rather than hardcoded secrets.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 28, 2026, 05:34 AM