streamlit-to-marimo
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run
uvx marimo checkto validate the resulting notebook. This involves the use of theuvpackage runner to execute themarimocommand-line interface. In the context of this skill, these are legitimate development tools associated with the vendor. - [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it requires the agent to read and process external Streamlit application code.
- Ingestion points: The agent reads the Streamlit app source code as specified in Step 1 of
SKILL.md. - Boundary markers: None explicitly defined in the instructions to delimit the external code from the agent's instructions.
- Capability inventory: The agent is instructed to create new files (Step 2) and execute shell commands via
uvx marimo check(Step 5). - Sanitization: There is no mention of sanitizing or escaping the content of the Streamlit app before processing.
Audit Metadata