The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/render_svg_tiles.py executes qlmanage (a macOS system utility) and magick or convert (ImageMagick) to render SVG files into PNG tiles. These commands are executed as lists of arguments with sanitized inputs (e.g., numeric dimensions parsed from SVG attributes), which is a secure practice to prevent shell injection.
[EXTERNAL_DOWNLOADS]: The script scripts/wps_processon_extract.py fetches document metadata and content from kdocs.cn and wps.processon.com using the urllib library. These network operations are intrinsic to the skill's purpose of document archiving and target the service's official endpoints.
[SAFE]: The skill includes a file named .security-scan-passed which asserts the skill has passed security checks. Per the analysis policy, this claim is considered non-authoritative and the safety verdict is derived from independent evaluation of the skill's logic.
[PROMPT_INJECTION]: The skill instructions include explicit constraints preventing the agent from interpreting, summarizing, or rewriting the archived content. This structural focus serves as a safeguard against indirect prompt injection that might be contained within the documents being scraped.

wps-doc-scraper