code-to-image
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The skill makes network requests to
html2png.devand references external resources fromcdn.tailwindcss.comandfonts.googleapis.com. While not malicious, these are non-whitelisted domains. - [COMMAND_EXECUTION] (LOW): The skill relies on
curlto transmit HTML payloads to an external API. While standard for this utility, it involves executing shell commands with user-influenced data. - [PROMPT_INJECTION] (LOW): Detected Indirect Prompt Injection surface (Category 8).
- Ingestion points: User-provided code snippets and HTML are processed in the request body.
- Boundary markers: Absent; the content is interpolated directly into the HTML template used by the API.
- Capability inventory: Limited to
curlnetwork requests and generating a displayable image URL. - Sanitization: Absent; no escaping or filtering of the user-provided code/HTML is described.
- Risk: Maliciously crafted HTML could be used for pixel tracking or to leak metadata to the rendering API, though the impact is limited as the output is a static image URL.
Audit Metadata