The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill employs several dynamic code execution patterns:
The script scripts/validate-presenter-mode.mjs uses the node:vm module's runInNewContext function to parse and evaluate the SPEAKER_NOTES array from the HTML files it validates.
The script scripts/validate-swiss-deck.mjs uses createRequire and dynamic import() to locate and load the playwright library from computed file system paths.
The assets/template.html file uses dynamic import() to load the Motion One library, falling back to a CDN if the local copy is unavailable.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest external user content (outlines, data, article links) and interpolate it into HTML templates. This creates a standard injection surface where instructions embedded in the source materials could influence the agent's behavior during the generation process.
[COMMAND_EXECUTION]: The SKILL.md instructions guide the agent to perform update checks and repository synchronization using git fetch and git pull commands within the skill's root directory, although it explicitly requires user confirmation.
[EXTERNAL_DOWNLOADS]: The generated presentations and templates reference resources from well-known services, including fonts from Google Fonts (fonts.googleapis.com), icons from Lucide (unpkg.com), mapping components from MapLibre (unpkg.com), and animation libraries from jsdelivr.net.

guizang-ppt-skill