creating-social-graphics
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches styling and font resources from well-known and trusted external services, specifically Tailwind CSS via cdn.tailwindcss.com and Google Fonts via fonts.googleapis.com.- [COMMAND_EXECUTION]: The skill instructs the agent to perform local command-line operations, including installing the playwright package through npm and executing a local Node.js rendering script located at scripts/render-to-png.js.- [DYNAMIC_EXECUTION]: The skill uses Playwright to dynamically render HTML and CSS templates into PNG images, which involves the use of a browser engine to process and execute code.- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes user-defined design content in a high-capability environment.
- Ingestion points: User-provided design content is interpolated directly into the tag of an HTML template in SKILL.md.
- Boundary markers: The template lacks explicit boundary markers or instructions to isolate user content from the rendering logic or to ignore embedded script tags.
- Capability inventory: The skill utilizes Playwright to render HTML (which can execute JavaScript) and has the capability to write the resulting image file to the local filesystem.
- Sanitization: No sanitization, filtering, or escaping is described for the user-provided HTML content before it is passed to the browser engine for rendering.
Audit Metadata