Finance Tracker
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious behavior detected. The skill instructions and provided code blocks focus on financial planning and analysis using standard methodologies.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process financial datasets which represent an ingestion surface for untrusted data. Evidence: (1) Ingestion points: SQL queries targeting financial_data and Python classes processing historical_data in SKILL.md. (2) Boundary markers: None explicitly defined in the provided templates. (3) Capability inventory: Limited to data processing and visualization; no tools for network access, file writing, or system command execution are invoked in the skill code. (4) Sanitization: No explicit sanitization or validation logic for external data inputs is implemented in the provided code snippets.
Audit Metadata