jupyter-to-marimo
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run marimo convert and marimo check using the uvx tool to transform file formats and validate the resulting output.
- [EXTERNAL_DOWNLOADS]: The uvx command dynamically fetches the marimo package from PyPI. This is a vendor-provided tool used for its intended purpose and matches the author's identity.
- [PROMPT_INJECTION]: The skill processes untrusted user-provided Jupyter notebooks, which presents a surface for indirect prompt injection if a notebook contains malicious instructions.
- Ingestion points: The agent reads converted Python files derived from untrusted .ipynb sources.
- Boundary markers: No specific delimiters are mentioned for the read operation.
- Capability inventory: The agent has access to command execution via uvx subprocesses.
- Sanitization: The process incorporates the marimo check command to validate and lint the generated code before final review.
Audit Metadata