xlsx
Pass
Audited by Gen Agent Trust Hub on Mar 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The helper script
recalc.pyexecutes system binaries to perform spreadsheet operations via LibreOffice.\n - Evidence: The script utilizes
subprocess.runto callsofficefor headless spreadsheet processing, as well astimeoutorgtimeoutfor process management.\n- [COMMAND_EXECUTION]: The scriptrecalc.pydynamically writes a LibreOffice StarBasic macro to the local file system to enable formula recalculation.\n - Evidence: The
setup_libreoffice_macrofunction creates or overwritesModule1.xbain the user's LibreOffice configuration directory.\n- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by ingesting data from external spreadsheet files without sanitization.\n - Ingestion points: Data is loaded from user-provided files using
pandas.read_excelandopenpyxl.load_workbookas described inSKILL.md.\n - Boundary markers: No delimiters or specific safety instructions are used to distinguish between spreadsheet data and agent instructions.\n
- Capability inventory: The agent has the ability to write files to the system and execute subprocesses through the provided recalculation script.\n
- Sanitization: No mechanisms are present to validate or sanitize cell content before it is processed by the agent.
Audit Metadata