check-understanding
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill retrieves curriculum metadata and lesson documentation from the author's GitHub repository (fancyboi999/ai-engineering-from-scratch-zh) to provide up-to-date quiz content. This represents expected behavior for a learning assistant based on the skill's primary purpose.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests text from external markdown files to generate quiz questions, which creates a theoretical surface for indirect prompt injection if those source files were to contain malicious instructions.\n
- Ingestion points: Remote GitHub repository files and local documentation paths defined in SKILL.md.\n
- Boundary markers: None identified in the prompt logic for processing external content.\n
- Capability inventory: Limited to local file reading and network GET requests to GitHub via URL fetching.\n
- Sanitization: No explicit content filtering or sanitization logic is described.
Audit Metadata