The Agent Skills Directory

[SAFE]: The skill performs legitimate image processing tasks using standard libraries (psd-tools, scikit-image). The core script scripts/export_slices.py focuses on parsing PSD structures and saving layers as PNG files on the local filesystem.
[COMMAND_EXECUTION]: The script uses argparse to handle user-provided command-line arguments for input/output paths. It creates directories using pathlib.Path.mkdir and writes files using the PIL (Pillow) image saving interface, which are standard operations for a file utility.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data in the form of PSD files and optional JSON mapping files. This presents an indirect injection surface where a malicious file could attempt to influence the skill's behavior.
Ingestion points: The skill reads PSD files and JSON mapping files provided via the --psd and --mapping-json arguments.
Boundary markers: The script does not use specific delimiters or instructions to ignore embedded commands in the processed data.
Capability inventory: The skill has the capability to write files and create directories on the local filesystem.
Sanitization: The script implements sanitize_filename which filters invalid characters (e.g., slashes, colons) from layer names to prevent directory traversal. However, filenames sourced directly from a user-provided JSON mapping file bypass this sanitization step, potentially allowing for path traversal if a malicious JSON is used.

psd-slicer