teaching-site

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill references the use of standard development tools, specifically npm for serving the static site and Playwright for automated UI verification and PDF generation.
  • [PROMPT_INJECTION]: The skill processes user-provided markdown files (e.g., overview.md, content.md) to generate site metadata and code, which constitutes an indirect prompt injection surface. Ingestion points: overview.md, day{n}/outline.md, and day{n}/content.md. Boundary markers: Uses HTML comments (e.g., <!-- thin-demo stub -->) for generated stubs. Capability inventory: Activation of sub-skills for code generation and local file modification. Sanitization: Relies on default agent handling; no explicit sanitization is described.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 01:17 PM
Security Audit — agent-trust-hub — teaching-site