design-qa-review-skill

Warn

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches source code from an untrusted GitHub repository located at https://github.com/ColourCloudSky/design-qa-review-skill.git.
  • [REMOTE_CODE_EXECUTION]: The installation instructions direct the user to git clone the repository and then execute npm install or pip install on the downloaded content, followed by running the design-qa-review command. This represents execution of unverifiable remote code.
  • [CREDENTIALS_UNSAFE]: The skill requires the configuration of sensitive environment variables, specifically FIGMA_ACCESS_TOKEN and BROWSERLESS_API_KEY, which are then used for network operations to external services.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: The skill processes data from arbitrary external URLs provided via --page-url and --mockup-url as well as local image files.
  • Boundary markers: There are no instructions or mechanisms described to distinguish between the visual data being reviewed and potential malicious instructions embedded within the UI/web content.
  • Capability inventory: The skill executes shell commands, performs network requests to Figma/Browserless, and writes reports to the file system.
  • Sanitization: There is no mention of sanitizing the content retrieved from external pages before processing it through the design comparison logic.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 16, 2026, 03:58 AM