figma-swiftui

Pass

Audited by Gen Agent Trust Hub on Oct 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill uses established platform tools (get_design_context, use_figma) and targets well-known Figma domains, following standard integration patterns for the figma vendor environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from Figma designs, creating a potential surface for indirect prompt injection. This is mitigated by the skill's detailed instructions on semantic pattern recognition and strict mapping to system components.\n
  • Ingestion points: Figma design context and properties (e.g., node names, labels) fetched via get_design_context in references/design-to-code.md.\n
  • Boundary markers: The skill uses structured tags like <CodeConnectSnippet> and <SFSymbol> to distinguish external data.\n
  • Capability inventory: The agent is authorized to generate and write SwiftUI code files to the project directory.\n
  • Sanitization: The skill provides logic for validating SF Symbol names, text styles, and system-defined component types before translation.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 15, 2026, 08:13 AM