moai
Pass
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes Python and Bash commands to validate the integrity and content of generated office documents. It executes automated checks using libraries like 'python-docx', 'python-pptx', and 'openpyxl' to ensure that the final output meets the user's structural requirements.
- [CREDENTIALS_UNSAFE]: During the initialization process, the skill prompts users to input API keys for legitimate services, including the Public Data Portal, KIPRIS, and Nano Banana. These credentials are stored in environment files or the agent's persistent memory to facilitate cross-project functionality.
- [PROMPT_INJECTION]: The system dynamically generates a project-specific 'CLAUDE.md' file that serves as an instruction set for the agent. This generation process follows structured templates and incorporates quality evaluation steps to manage the persona and routing logic without overriding base safety guidelines.
Audit Metadata