curriculum-package-pdf
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- Prompt Injection (SAFE): The skill documentation and parameters contain no instructions intended to bypass AI safety guardrails or extract system prompts.
- Data Exposure & Exfiltration (SAFE): No hardcoded secrets, sensitive file paths, or network exfiltration vectors were identified.
- Remote Code Execution (SAFE): There are no remote script downloads or dynamic execution patterns such as eval() or exec().
- Command Execution (SAFE): Although the documentation describes CLI usage, the Python implementation does not use subprocess or shell execution modules.
- Indirect Prompt Injection (SAFE): The skill accepts a content path but does not process the file's contents, eliminating the risk of data-driven injection. Evidence: 1. Ingestion points: content_path parameter. 2. Boundary markers: Absent. 3. Capability inventory: None (returns dictionary data only). 4. Sanitization: Absent.
Audit Metadata