The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses several shell scripts (render.sh, fetch_image.sh, booklet.sh, printable.sh, print_booklet.sh) to perform its duties. These scripts invoke system commands such as curl, lp, python, pdftoppm, and headless Google Chrome.
[COMMAND_EXECUTION]: The scripts/make_printable.py script uses subprocess.run to call pdftoppm for rasterizing PDF pages. This is a functional requirement for the skill's purpose of ensuring print compatibility and does not use untrusted input in a way that enables injection.
[EXTERNAL_DOWNLOADS]: scripts/fetch_image.sh uses curl to download images from user-provided URLs. It includes a browser User-Agent and referer to avoid blocking by CDNs. The skill instructions explicitly tell the agent to 'Read each downloaded image to confirm it's the right subject' and not a 404 or logo before use.
[DATA_EXPOSURE]: The skill instructions mention reading ~/.claude/PROSE.md and invoking a stop-slop skill to audit writing quality. This accesses local configuration files intended for agent behavior customization, which is standard for this platform environment.

magazine-explainer