redesign-existing-projects
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to scan existing codebases and apply modifications to improve UI/UX. This workflow exposes the agent to potentially malicious instructions embedded within the project files (e.g., in source code comments or documentation) that could influence its actions during the modification phase.\n
- Ingestion points: External project codebase (accessed via the 'Scan' step in
SKILL.md)\n - Boundary markers: Absent\n
- Capability inventory: File modification and code editing (via the 'Fix' step in
SKILL.md)\n - Sanitization: Absent\n- [EXTERNAL_DOWNLOADS]: The skill references picsum.photos as a source for placeholder background and hero imagery when local assets are unavailable.
Audit Metadata