soia-dev-officecli-ops
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill initiates subprocesses to run the
officecliexecutable. These calls are implemented using list-based arguments, which prevents shell command injection.\n- [EXTERNAL_DOWNLOADS]: The documentation directs users to the GitHub repository forOfficeCLI(iOfficeAI/OfficeCLI) for manual installation. No automated downloading or execution of remote scripts is performed by the skill itself.\n- [PROMPT_INJECTION]: Vulnerability to indirect prompt injection through the processing of external Office documents.\n - Ingestion points: Input files provided to
scripts/officecli_inspect.pyandscripts/officecli_safe.py.\n - Boundary markers: Absent; there are no specific instructions to the agent to ignore embedded commands within the files.\n
- Capability inventory: File reading, structural modification, and schema validation via the
officeclitool as seen inscripts/officecli_inspect.pyandscripts/officecli_safe.py.\n - Sanitization: Subprocess calls use argument lists rather than shell strings, providing protection against typical injection attacks at the command line level.
Audit Metadata