redesign-existing-projects
Pass
Audited by Gen Agent Trust Hub on Jul 3, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The skill provides comprehensive design guidelines that follow industry best practices for typography, accessibility, and user interface consistency. It emphasizes non-destructive edits and focuses on enhancing existing tech stacks rather than performing risky rewrites.
- [PROMPT_INJECTION]: The skill processes untrusted data by scanning the existing codebase of the project it is intended to redesign. This creates an indirect prompt injection surface where malicious instructions embedded in the project files (such as in comments or text content) could potentially influence the AI's behavior.
- Ingestion points: codebase scanning defined in SKILL.md.
- Boundary markers: Absent; there are no instructions to ignore or delimit embedded commands within the scanned code.
- Capability inventory: The skill can modify project files and write documentation to the 'llmwiki/' directory.
- Sanitization: Absent; the skill does not specify any validation or filtering of the content being read.
- [EXTERNAL_DOWNLOADS]: Fetches placeholder imagery from the well-known service 'picsum.photos' when real project assets are missing.
Audit Metadata