budget-forecast
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: Comprehensive analysis of the instructions, scripts, and provided datasets reveals no malicious behavior, prompt injection, or data exfiltration attempts.
- [EXTERNAL_DOWNLOADS]: The skill references established financial education and news platforms including Investopedia, Harvard Business Review (HBR), and Corporate Finance Institute (CFI).
- [COMMAND_EXECUTION]: The skill includes a local utility script (scripts/calculate.py) for accurate financial modeling; the script is limited to mathematical operations and lacks network or sensitive file access capabilities.
- [SAFE]: While the skill ingests external data (CSV) and user input, the risk of indirect prompt injection is mitigated by the restricted mathematical nature of the calculations. Evidence: Ingestion points: SKILL.md and history_and_drivers.csv; Boundary markers: None; Capability inventory: local script execution for math; Sanitization: None.
Audit Metadata