design-system

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches stock images from Pexels, a well-known service, for use as slide backgrounds. It also references font and script assets from Google Fonts and jsDelivr's CDN. These are documented as standard design assets.
  • [COMMAND_EXECUTION]: The slide-token-validator.py script executes a local sibling script, html-token-validator.py, using subprocess.run. This execution is restricted to the local Python environment for the purpose of validating design compliance.
  • [INDIRECT_PROMPT_INJECTION]: The slide generation process ingests user-provided text and JSON configuration, creating a surface for indirect injection.
  • Ingestion points: User commands for slide creation and local design-tokens.json files.
  • Boundary markers: The skill implements explicit HTML escaping for all user-supplied fields.
  • Capability inventory: Includes reading/writing local project files and executing local validation scripts.
  • Sanitization: Comprehensive sanitization is present in generate-slide.py via the _e() helper for HTML escaping and _safe_url() for protocol verification (allowing only http/https/relative links).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 10:36 AM