skills/ghaida/intent/journey/Gen Agent Trust Hub

journey

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses Model Context Protocol (MCP) tools, such as mcp__claude_ai_Figma__use_figma and mcp__pencil__batch_design, to facilitate diagram creation in external design applications. These are intended uses of the platform's tool capabilities.
  • [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by interpolating user-supplied data into its visual outputs.
  • Ingestion points: User descriptions of design flows provided at runtime.
  • Boundary markers: The generated HTML artifacts are bounded by markdown code blocks.
  • Capability inventory: The skill leverages authorized MCP design tools and static code generation.
  • Sanitization: There is no explicit sanitization for user-provided labels, though the risk is minimal given the static and visual nature of the output.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 04:34 AM
Security Audit — agent-trust-hub — journey