aliyun-qwen-ocr

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a legitimate tool for interacting with Alibaba Cloud OCR services. All components, including the Python helper script and documentation, align with the stated goal of document and image text extraction.\n- [PROMPT_INJECTION]: The skill handles external data inputs (images and prompt strings) which are subsequently processed by an OCR model. This represents an indirect prompt injection surface typical of multimodal processing tools.\n
  • Ingestion points: User-supplied image URLs and prompts are accepted by the prepare_ocr_request.py script in the image and prompt arguments.\n
  • Boundary markers: The skill relies on the standard JSON structure of the API request and does not explicitly wrap untrusted content in additional safety delimiters or instructions.\n
  • Capability inventory: The script includes functionality to write JSON request files to the local file system via the output argument.\n
  • Sanitization: The script performs standard JSON serialization of input parameters without specific filtering or sanitization of the image content or prompt text.\n- [EXTERNAL_DOWNLOADS]: The instructions recommend installing the requests library from the Python Package Index (PyPI). This is a well-known and standard service, and the dependency is appropriate for the skill's intended functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 02:27 AM