digital-college-yearbook

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill contains a surface for indirect prompt injection through its handling of external student data files.
  • Ingestion points: The skill fetches data from data/students.json and other related files within the roster.js implementation.
  • Boundary markers: No delimiters or ignore instructions are provided to the agent to distinguish between student-provided data and executable code in the generated output.
  • Capability inventory: The skill generates code that utilizes innerHTML for DOM manipulation based on external strings such as student names and quotes.
  • Sanitization: There is no logic included to escape or sanitize the data before it is interpolated into the generated website's DOM.
  • [EXTERNAL_DOWNLOADS]: The skill downloads utility and animation libraries from established content delivery networks.
  • Evidence: The SKILL.md file specifies the use of GSAP and Fuse.js from cdnjs.cloudflare.com.
  • Note: These downloads target well-known services and are used for standard functional purposes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 01:05 AM
Security Audit — agent-trust-hub — digital-college-yearbook