qwen-edit
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill documentation is focused entirely on legitimate image editing workflows using the Qwen-Image-Edit model via a CLI tool.
- [COMMAND_EXECUTION]: The skill instructions utilize a local CLI wrapper (
tools/image_edit.py) to perform operations. The usage patterns shown inSKILL.mdandparameters.mdinvolve standard command-line argument passing for input files and model parameters (steps, guidance, etc.). - [PROMPT_INJECTION]: Instructions labeled as 'IMPORTANT' or 'Critical' within the documents are strictly related to improving image generation quality and preventing visual artifacts (e.g., negative prompts for avoiding blurry results). There are no attempts to bypass agent safety filters or override system prompts.
- [DATA_EXPOSURE]: The skill interacts with local image files (e.g.,
photo.jpg) provided by the user for processing. No access to sensitive system paths, credentials, or environment variables was observed.
Audit Metadata