skills/eva813/skills-base/ai-pm/Gen Agent Trust Hub

ai-pm

Pass

Audited by Gen Agent Trust Hub on Aug 16, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data from Figma designs, which constitutes a potential injection surface.
  • Ingestion points: The skill accepts a user-provided Figma URL or Node ID as its primary input and uses the mcp-figma tool to extract data.
  • Boundary markers: The instructions do not specify explicit boundary markers or "ignore instructions" warnings when processing text content extracted from Figma nodes (such as component names or labels).
  • Capability inventory: The skill invokes the mcp-figma tool, writes a draft-spec.md file to the local environment, and generates JSON handoff payloads.
  • Sanitization: No specific sanitization or filtering logic is mentioned for the data retrieved from the Figma API.
  • Context: While this creates a surface for indirect injection if a design contains malicious text, the risk is limited to the content of the generated documentation and does not include execution of arbitrary code.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 16, 2026, 12:30 AM
Security Audit — agent-trust-hub — ai-pm