ui-prototype-preview
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted content from brainstorm files (
brain-storm/**/*.md) and interpolates them into HTML templates without sanitization. This creates a surface for indirect prompt injection where instructions hidden in those files could influence the agent's generation behavior or embed malicious scripts in the resulting prototypes.\n - Ingestion points: Brainstorm idea files located in the
brain-storm/directory.\n - Boundary markers: Absent; no delimiters or system warnings are used to isolate content from the brainstorm files.\n
- Capability inventory:
Write(to save HTML),Bash(to open files), andAskUserQuestiontools are used.\n - Sanitization: Absent; the skill lacks mechanisms to sanitize or validate external content before it is placed into HTML placeholders like
{{BODY}}or{{FONT_PRELOADS}}.\n- [COMMAND_EXECUTION]: The skill utilizes theBashtool to optionally open generated prototype HTML files in a local browser for previewing. While this is a common developer task, executing shell commands based on generated file paths requires strict adherence to naming constraints to avoid command injection risks.
Audit Metadata