tax-season-organizer

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes financial data ingested from external CSV/Excel files and accounting software connectors. This ingestion of untrusted data represents a potential surface for indirect prompt injection attacks.
  • Ingestion points: Data is ingested via the ~~бухгалтерия and ~~платежи MCP connectors, as well as manual CSV/Excel file uploads for fallback processing.
  • Boundary markers: The skill includes mandatory instructions to preface all output with a disclaimer stating the material is for accountant review only and is not a tax consultation. It also requires the listing of all assumptions made during calculation.
  • Capability inventory: The skill does not possess high-risk capabilities such as arbitrary shell command execution, file system modification, or external network requests to non-whitelisted domains.
  • Sanitization: The skill relies on standard platform guardrails; no specific input validation or escaping logic for the CSV data is implemented within the prompt instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 04:25 PM
Security Audit — agent-trust-hub — tax-season-organizer