de-div
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use standard CLI utilities like
grepto scan the project files for specific HTML tags (<div>,<span>) and attributes (role=) to evaluate the codebase's current semantic state. - [EXTERNAL_DOWNLOADS]: The skill references authoritative web standards documentation from W3C and MDN, as well as a collection of HTML recipes hosted on the author's personal domain (
tonyalicea.dev), which is considered a legitimate vendor resource. - [INDIRECT_PROMPT_INJECTION]: Because the skill processes user-supplied code (HTML, JSX, Vue templates), it has an attack surface where malicious instructions could be embedded in the code being analyzed. However, the skill mitigates this by using a structured decision-making workflow that requires user confirmation for ambiguous cases and produces a final report for review.
Audit Metadata