learning-global-accessibility

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill is designed to process external course content for accessibility validation, which introduces a surface for indirect prompt injection attacks.
  • Ingestion points: The source_content parameter in skill.py accepts data from potentially untrusted sources (e.g., external course files).
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the code or prompt templates.
  • Capability inventory: The current code returns a JSON report and references artifacts; the broader system may use these outputs for automated decision-making.
  • Sanitization: The execute method lacks input validation or sanitization for the content being processed.
  • [Code Quality] (SAFE): The skill.py file contains a functional error (NameError) because the variable skill_dir is referenced in the execute method but never defined. This will cause the skill to fail at runtime but does not represent a security vulnerability.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:15 PM