invoice-organizer
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill functions as a local utility for bookkeeping tasks, processing CSV files to categorize expenses and identify potential duplicates.
- [SAFE]: Analysis of the Python script and instructions confirmed the absence of network connectivity, credential harvesting, or unauthorized file access.
- [SAFE]: The skill relies on standard Python functionality and does not employ dynamic code execution or privilege escalation techniques.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied CSV data, which represents a theoretical attack surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the context via the receipts.csv file processed by invoice_categorizer.py.
- Boundary markers: The script's output lacks explicit delimiters to distinguish processed data from instructions for the agent.
- Capability inventory: The script possesses no high-risk capabilities such as network operations, file writing, or shell command execution.
- Sanitization: While numeric data is filtered in the parse_amount function, the vendor and description fields are interpolated into the output without sanitization in the render_human function.
Audit Metadata