The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and process external Western blot images and data, which constitutes a potential attack surface for indirect instructions.
Ingestion points: The skill utilizes tools such as analyze_pixel_distribution and find_roi_from_image to ingest external image files and extract band intensity data.
Boundary markers: There are no explicit instructions to the agent to ignore or delimit embedded text or metadata that might be present in the external images or raw data files.
Capability inventory: The skill possesses the capability to write to the file system to save verification images (wb_grid_verification.png), visualization plots (psmad2_quantification.png), and quantification reports (CSV/Excel).
Sanitization: The instructions do not define validation or sanitization routines for the data extracted from the external image sources before it is used to generate final reports or visualizations.

western-blot-quantification