import-transactions

Pass

Audited by Gen Agent Trust Hub on Apr 14, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests untrusted transaction descriptions from bank files, creating a surface for indirect prompt injection. This is evaluated as safe given the data is used for matching and categorization within a limited financial context. Evidence: Ingestion points: Bank export files (CSV, OFX, QIF) specified in SKILL.md. Boundary markers: Absent. Capability inventory: transaction_search and categorize tools. Sanitization: Not specified.
  • [EXTERNAL_DOWNLOADS]: The documentation includes informational links to official banking websites and a transaction conversion utility for manual user workflows. These references are for user guidance and do not involve automated network downloads or remote code execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 14, 2026, 01:50 PM