resume-to-personal-site

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill workflow requires the agent to download and install several well-known libraries from the npm registry.
  • UI and animation libraries include framer-motion and lucide-react.
  • Atmospheric visual effects use pixi.js.
  • PDF export and page capture functionality relies on html2canvas and jsPDF.
  • [REMOTE_CODE_EXECUTION]: The skill generates and executes a custom script for runtime asset management.
  • The agent is instructed to create a TypeScript file named scripts/generate-assets.ts to handle calls to the OpenAI Images API.
  • This generated script is subsequently executed in the local environment using the tsx runner to populate the project's asset folder.
  • [COMMAND_EXECUTION]: Project setup and execution involve running multiple shell commands.
  • Standard package management and build commands such as npm install and npm run build are used.
  • The custom asset generation script is executed via a command-line interface.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface due to the ingestion of untrusted resume data.
  • Resume content (text, profiles, or PDF extracts) is directly interpolated into prompts for generating website copy, positioning, and image descriptions.
  • While the skill includes comprehensive 'Anti-AI' rules to maintain a specific tone, it lacks explicit data boundary markers or escaping for the ingested resume content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 01:53 AM
Security Audit — agent-trust-hub — resume-to-personal-site