Skills
skills/ericlovold/skills/render-check/Gen Agent Trust Hub

render-check

Fail

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use sudo -u postgres to manage the PostgreSQL database service, which requires granting the agent elevated system-level permissions.
  • [COMMAND_EXECUTION]: The verification ritual utilizes node -e to execute a dynamic JavaScript payload for headless browser automation via Playwright.
  • [EXTERNAL_DOWNLOADS]: The command npx next dev downloads and executes packages from the public npm registry at runtime.
  • [PROMPT_INJECTION]: The skill creates an indirect prompt injection surface by rendering local application pages and instructing the agent to process the visual output. Maliciously crafted data in the application UI (e.g., HTML comments or text content) could influence the agent's reasoning.
  • Ingestion points: Headless browser rendering http://localhost:3112/<page> (SKILL.md).
  • Boundary markers: None; the agent is instructed to read the screenshots directly.
  • Capability inventory: sudo, node -e, npx, and curl (SKILL.md).
  • Sanitization: None; the agent processes the raw visual output of the rendered page.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jul 12, 2026, 08:13 PM
Security Audit — agent-trust-hub — render-check