iblai-api-studio
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of user-provided content (HTML fragments, PDFs, and problem definitions) to be sent to external APIs, creating a potential surface for indirect prompt injection.
- Ingestion points: User-supplied course outlines, HTML blocks, and quiz/problem definitions are processed in SKILL.md.
- Boundary markers: No explicit delimiter or "ignore instructions" markers are defined in this high-level guide.
- Capability inventory: The skill suite possesses network operation capabilities (POST requests to ibl.ai domains) and local file access (reading studio.env).
- Sanitization: Not explicitly addressed in this guide; reliance is placed on specific sub-skills and platform-level validation.
Audit Metadata