teach-me-anything

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill establishes a local tutoring environment. All file operations and state management are conducted within the designated workspace directory.
  • [SAFE]: The workspace uses static local assets provided with the skill, including a custom Markdown renderer (md.js). No external scripts or dependencies are downloaded or executed from remote sources.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection due to its core function of ingesting and rendering external educational material.
  • Ingestion points: The agent incorporates external sources into local HTML and Markdown files (sources.html, lessons, glossary).
  • Boundary markers: No explicit markers are defined in the templates to delimit external content.
  • Capability inventory: The agent writes and updates files within the local workspace directory.
  • Sanitization: The assets/md.js script performs basic escaping of HTML characters but lacks comprehensive URL sanitization.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 06:30 AM
Security Audit — agent-trust-hub — teach-me-anything