classify-transactions

Pass

Audited by Gen Agent Trust Hub on Jul 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script scripts/classify_transactions.py and uses npx to run the @cynco/accounting-skills utility. These commands are necessary for the primary accounting functions of the skill and are restricted to the provided transaction datasets.
  • [EXTERNAL_DOWNLOADS]: The use of npx @cynco/accounting-skills involves fetching the package from the NPM registry. This dependency is authorized as it belongs to the vendor 'cynco-labs' and is a standard delivery method for such tools.
  • [PROMPT_INJECTION]: The skill processes external financial data from workpapers/transactions.json, which presents a surface for indirect prompt injection. The risk is minimized by the skill's reliance on structured script logic and specific accounting standards checklists.
  • Ingestion points: workpapers/transactions.json containing bank transaction descriptions
  • Boundary markers: No specific delimiters are defined in the instructions for the data being analyzed
  • Capability inventory: Execution of local Python and Node.js scripts with file system write access
  • Sanitization: The instructions do not specify sanitization or escaping of transaction data before analysis
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 9, 2026, 02:13 PM
Security Audit — agent-trust-hub — classify-transactions