redesign
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run the
npx ui-craft-detectcommand against user-provided paths or URLs. This executes a CLI tool to audit the current state of a site or application. - [EXTERNAL_DOWNLOADS]: The use of
npxinvolves downloading and executing theui-craft-detectpackage. This is a resource provided by the skill vendor. - [PROMPT_INJECTION]: The skill processes untrusted data from external live pages, creating a vulnerability to indirect prompt injection where content on a website could influence the agent's code-generation and editing behavior.
- Ingestion points: External URLs audited via
npx ui-craft-detect <url>in SKILL.md. - Boundary markers: None identified to isolate audited data from the agent's instruction context.
- Capability inventory: Direct modification of project source code and layout files.
- Sanitization: No validation or sanitization of the audited content is described.
Audit Metadata