lbo-model
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions include a verification step that executes a local Python script (
python /mnt/skills/public/xlsx/recalc.py model.xlsx) to validate Excel formulas. This is a functional requirement for the skill's purpose and uses a hardcoded path and filename, preventing command injection. - [INDIRECT_PROMPT_INJECTION]: The skill acts as an ingestion surface for external data by processing user-provided Excel templates.
- Ingestion points: User-attached Excel templates and the local
examples/LBO_Model.xlsxfile. - Boundary markers: None identified; the skill directly processes the structure and content of provided Excel files.
- Capability inventory: The skill writes Excel formulas and executes the
/mnt/skills/public/xlsx/recalc.pyscript for validation. - Sanitization: The skill relies on the
recalc.pyscript to validate formula correctness but does not explicitly filter for prompt-based instructions hidden within Excel cells.
Audit Metadata