create-visual-card
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads font resources from Google Fonts' official CDN (fonts.googleapis.com) to style the visual cards.
- [EXTERNAL_DOWNLOADS]: The skill requires the Playwright browser automation library and may trigger the download of the Chromium browser binary using
npx playwright install chromium. - [COMMAND_EXECUTION]: The skill executes a local Node.js script
scripts/screenshot.mjsto perform the screenshot operation. - [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to the way user content is handled.
- Ingestion points: User-supplied text and data are used as the primary content for the generated HTML cards.
- Boundary markers: The instructions do not define clear boundaries or 'ignore' directives for the interpolated user content.
- Capability inventory: The skill has the ability to write files (HTML and PNG) and execute a headless browser (Playwright) via a local script.
- Sanitization: There is no instruction to sanitize, escape, or validate the user-provided content before it is embedded into the HTML template. This could allow a malicious user to inject scripts or HTML elements that execute within the context of the headless browser during the screenshot process.
Audit Metadata