jupyter-notebook-skills
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION] (LOW): The templates provide a surface for indirect prompt injection by ingesting untrusted external data. Nested Evidence:
- Ingestion points: All three templates (eda_template.md, ml_template.md, visualization_template.md) use 'pd.read_csv("your_data.csv")' to load data.
- Boundary markers: None identified; untrusted data is processed directly without delimiters or guardrails.
- Capability inventory: The templates include file-write capabilities such as 'joblib.dump' for saving models and 'plt.savefig' for saving visualizations.
- Sanitization: No input validation or sanitization of the CSV content is performed before processing or displaying findings.
- [SAFE] (SAFE): No evidence of hardcoded credentials, remote code execution, or obfuscation was found in the provided templates.
Audit Metadata