The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches external assets from well-known and trusted providers. Background images in scripts/fetch-background.py are sourced from images.pexels.com, and brand fonts in scripts/generate-slide.py are loaded from fonts.googleapis.com. Additionally, SKILL.md references the Chart.js library via the cdn.jsdelivr.net content delivery network. These are standard development practices and present no security risk.
[COMMAND_EXECUTION]: The script scripts/slide-token-validator.py uses subprocess.run to call an internal utility script, html-token-validator.py. This delegation is handled safely by passing arguments as a list rather than using a shell environment, which prevents command injection vulnerabilities.
[PROMPT_INJECTION]: The skill processes user-supplied data for token configuration and slide generation. The scripts/generate-slide.py file includes proactive security measures, such as HTML escaping of content and strict protocol validation for URLs (allowing only http, https, and local anchors). These measures effectively mitigate risks associated with indirect prompt injection or cross-site scripting (XSS) in the generated HTML documents.
[SAFE]: While several scripts (scripts/generate-slide.py, scripts/fetch-background.py) employ deep parent directory traversal logic to locate the project's asset folder, this behavior is used solely for accessing internal design resources. It does not target sensitive system files or credentials.

design-system