coo-advisor
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of operational frameworks and analytical tools for organizational leadership. No malicious behavior or security vulnerabilities were identified across the instructions or scripts.
- [COMMAND_EXECUTION]: Included scripts
okr_tracker.pyandops_efficiency_analyzer.pyare designed for data processing and business analysis. They utilize standard Python libraries (such as json and argparse) to read user-provided data and generate textual reports, with no evidence of unauthorized system access or arbitrary code execution. - [INDIRECT_PROMPT_INJECTION]: The skill features a data ingestion surface through scripts that process external JSON files. 1. Ingestion points: The
--inputflag in the included Python scripts. 2. Boundary markers: Absent within the scripts. 3. Capability inventory: Restricted to local file operations with no network access. 4. Sanitization: Relies on standard JSON parsing. The risk is deemed safe as the processing is focused on mathematical metrics and structured reporting rather than instruction interpretation.
Audit Metadata