alicloud-ai-entry-modelstudio-test
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFECREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill mentions the path
~/.alibabacloud/credentialsfor API authentication. This is the standard location for Alibaba Cloud credentials and is provided as guidance for user configuration.- [EXTERNAL_DOWNLOADS]: The skill installs thedashscopelibrary via pip. This is the official Python SDK for Alibaba Cloud's Model Studio (DashScope), which is a well-known and trusted service.- [COMMAND_EXECUTION]: The skill executes shell commands to manage a Python virtual environment and create directories for output artifacts. These operations are standard administrative tasks for the skill's testing lifecycle.- [PROMPT_INJECTION]: An indirect prompt injection surface is present. 1. Ingestion points: The skill readsSKILL.mdfiles from subdirectories in the repository. 2. Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified. 3. Capability inventory: The skill executes scripts and writes results to the local filesystem. 4. Sanitization: No sanitization of the content from the referencedSKILL.mdfiles is performed before script execution. This behavior is expected for a test runner skill.
Audit Metadata