design-email

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and audits existing project files, which constitutes a surface for indirect prompt injection.
  • Ingestion points: Reads package.json, .env files, and existing templates in emails/ or src/emails/ directories.
  • Boundary markers: Absent. There are no explicit instructions to the agent to disregard instructions embedded within the audited files.
  • Capability inventory: Executes shell commands for package installation (npm install), testing (npx), and function deployment (supabase functions deploy).
  • Sanitization: Absent. The skill reviews file content for quality and design without specific sanitization of potential embedded directives.
  • [EXTERNAL_DOWNLOADS]: The skill performs downloads from well-known and expected sources for its development purpose.
  • Fetches Node.js packages from the NPM registry.
  • Conducts web searches using Firecrawl for best practices research.
  • Resolves documentation IDs for the React Email library.
  • [COMMAND_EXECUTION]: The skill instructs the agent to run shell commands to manage the development environment.
  • Installs dependencies using npm install.
  • Deploys edge functions using the Supabase CLI.
  • Runs local dev servers and automated tests using npx email dev and Playwright.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 05:55 AM
Security Audit — agent-trust-hub — design-email