accounting-data-processor
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data including OCR results from uploaded photos and content from Excel/CSV files, creating an attack surface for embedded instructions.
- Ingestion points: Processes photo/OCR data and accounting files (Excel/CSV) as described in the instructions.
- Boundary markers: The skill lacks explicit boundary markers or instructions to the agent to disregard instructions embedded within the processed financial data.
- Capability inventory: The agent has the capability to write detailed logs to the file system (audit_logs/ directory) and interact with desktop accounting software via the desktop-rpa-computer-use tool.
- Sanitization: There is no evidence of sanitization, filtering, or validation of the extracted OCR text or file inputs before they are used in reconciliation logic or written to audit logs.
- [DATA_EXPOSURE]: The skill is instructed to log sensitive information, including Tax IDs (NPWP) and transaction summaries, to local files (Excel and CSV). While these are standard audit practices, they represent the storage of PII/sensitive data that should be protected by the host environment.
Audit Metadata