alibabacloud-pts-ops
Pass
Audited by Gen Agent Trust Hub on May 2, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides an interface for processing user-supplied data (such as target URLs, scene names, and configuration parameters) which are interpolated into shell commands.
- Ingestion points: Workflow definitions in
SKILL.mdand parameter tables. - Boundary markers: The skill uses structured JSON payloads for complex parameters, providing a degree of separation between data and command structure.
- Capability inventory: The skill executes Alibaba Cloud CLI (
aliyun) commands to manage cloud resources. - Sanitization: The skill includes mandatory instructions for the agent to confirm all user-customizable parameters before execution, serving as a human-in-the-loop safety checkpoint.
Audit Metadata