create-visualizer
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill utilizes the
$ARGUMENTSvariable to include user-supplied text directly in its operational instructions. - Ingestion points: The
$ARGUMENTSvariable inSKILL.mdis replaced with the user-provided environment name at runtime. - Boundary markers: The instructions lack delimiters or specific isolation markers to separate the untrusted input from the rest of the prompt logic.
- Capability inventory: The skill is designed to manage web development tasks, including generating Vite and TypeScript frontend code and handling replay playback.
- Sanitization: There is no evidence of filtering, escaping, or validation of the input, which could allow a malicious user to supply instructions that override the agent's behavior.
Audit Metadata