alicloud-ai-image-zimage-turbo
Warn
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: MEDIUMDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill accesses sensitive local files to retrieve authentication credentials for Alibaba Cloud services. Specifically,
scripts/generate_image.pyreads~/.alibabacloud/credentialsto extract API keys and parses.envfiles in the working directory and repository root for environment variables. - [PROMPT_INJECTION]: The skill processes untrusted user input which creates an attack surface for indirect prompt injection. 1) Ingestion points: User prompts are loaded from command-line arguments or local JSON files in
scripts/generate_image.py. 2) Boundary markers: No delimiters or protective instructions are implemented to isolate the user prompt from the system instructions when sent to the DashScope API. 3) Capability inventory: The skill can make network POST requests to the DashScope API, perform GET requests to download images from arbitrary URLs returned by the API, and write files to the local filesystem. 4) Sanitization: No input validation or sanitization is performed on the user-provided prompt before it is processed.
Audit Metadata