ian-xiaohei-svg-motion
Pass
Audited by Gen Agent Trust Hub on Jun 28, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting untrusted user content (such as articles or storyboard scenes) and interpolating it into generated HTML/SVG files.
- Ingestion points: User-provided Chinese articles and storyboard scripts as described in the workflow section of
SKILL.md. - Boundary markers: The skill does not provide instructions to use delimiters or ignore-embedded-instructions warnings when handling this data.
- Capability inventory: The skill generates HTML files with embedded Javascript execution capabilities via the GSAP library.
- Sanitization: There are no explicit instructions for the agent to sanitize or escape user-provided content before inserting it into the
index.htmltemplate, which could lead to cross-site scripting (XSS) or logic bypass if the generated files are opened in a browser.
Audit Metadata