marimo-notebook
Pass
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill provides standard instructions for interacting with the
marimoanduvCLI tools. These commands are typical for the development environment and are used for running, editing, and checking the integrity of notebooks. - [EXTERNAL_DOWNLOADS]: Documents the installation of necessary Python dependencies and browser binaries (via Playwright) from official and well-known registries to support features like PDF export and testing.
- [DATA_EXFILTRATION]: Includes explicit safety instructions in the
ANYWIDGET.mdreference file, directing the agent not to access sensitive local files (such as SSH keys or environment variables) when generating custom UI components. - [PROMPT_INJECTION]: While the skill illustrates the use of SQL and Javascript interpolation, it does so within the context of standard library features. There are no patterns suggesting attempts to bypass agent security filters or override system instructions.
Audit Metadata