The Agent Skills Directory

[SAFE]: The skill provides documentation and legitimate Python utility functions for scientific image validation. The logic focuses on verifying resolution (DPI), file formats, and color modes as required by AACR guidelines.
[DATA_EXPOSURE_AND_EXFILTRATION]: The Python scripts use the Pillow library to open local image files for metadata inspection. This is a standard operation for the skill's stated purpose, and no network exfiltration patterns or hardcoded credentials were observed.
[INDIRECT_PROMPT_INJECTION]: The skill contains a vulnerability surface for processing external data.
Ingestion points: The image_path parameter in the validate_cancer_res_figure and check_cancer_res_resolution functions in SKILL.md allows the agent to process external image files.
Boundary markers: None present; the skill assumes the provided path is for a valid image.
Capability inventory: File system read access via PIL.Image.open and console output via print.
Sanitization: Basic verification of image format against an accepted list.

cancer-research-figure-guide