financial-statement-analyzer

Pass

Audited by Gen Agent Trust Hub on Nov 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: Comprehensive analysis of the instructions and the provided Python script reveals no evidence of malicious intent, credential theft, or unauthorized system access.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted financial data from external reports provided by users.
  • Ingestion points: Data enters the agent's context when users provide financial statements (PDFs, images, or tables) to be processed as specified in SKILL.md.
  • Boundary markers: The skill uses a structured JSON schema to encapsulate extracted data before passing it to the analyzer script, creating a separation between raw input and processing logic.
  • Capability inventory: The associated script (scripts/analyze_financials.py) is limited to mathematical computation and reporting. It lacks capabilities for network exfiltration, arbitrary command execution, or persistent storage modification.
  • Sanitization: The script includes explicit data validation functions that verify input structure, data types, and array consistency.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 22, 2026, 01:11 AM