exam-ready
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data (PDFs, pasted notes, and syllabus text) which could contain hidden instructions designed to manipulate agent behavior.
- Ingestion points: As described in
SKILL.md, the skill accepts study material and syllabus topics provided by the user. - Boundary markers: The instructions include functional constraints such as 'stay strictly inside the supplied material' and 'Decline outside knowledge', alongside a rigid output template to guide the response.
- Capability inventory: The skill focus is limited to generating text-based study guides; it does not request tools for network access, file system writing, or arbitrary code execution.
- Sanitization: The skill relies on instructional prompts to ignore external knowledge but does not specify technical sanitization or content validation for the input data.
Audit Metadata