plan-uiux-unification

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill follows a non-destructive audit approach. It explicitly requires the agent to acknowledge a "Preservation Contract" that forbids unauthorized code removal, fabrication of data, or breaking changes to business logic.
  • [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface as it ingests untrusted data from the local codebase (via glob and grep) and external design research (via Firecrawl).
  • Ingestion points: Reads project files (e.g., tailwind.config.*, React/Vue components) and external web content via Firecrawl research.
  • Boundary markers: The preservation-contract.md and "Self-critique" sections act as instructional boundaries to maintain focus on the audit task and prevent obedience to embedded instructions.
  • Capability inventory: Limited to file reading and network research; code writing and tool execution for modification are explicitly gated until approved.
  • Sanitization: None explicitly defined for data interpolation, but instructions emphasize citing real paths and using existing content verbatim.
  • [EXTERNAL_DOWNLOADS]: The skill uses Firecrawl to fetch current best practices for design systems and accessibility. This is a legitimate functional requirement for a design audit and does not involve executing remote code.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 11:27 PM