sagemaker-to-zenml-migration
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided SageMaker pipeline code to generate a new ZenML project, creating a surface where malicious instructions in the input could influence the agent's behavior or output.
- Ingestion points: Processes SageMaker Python files, evaluation code, and infrastructure snippets provided by the user in Phase 1.
- Capability inventory: The agent is instructed to create directory structures, write multiple Python scripts, generate configuration YAMLs, and use CLI tools like GitHub (gh) for issue creation.
- Sanitization: There are no explicit instructions to sanitize user input or use boundary markers to prevent the agent from following instructions embedded within the source code being migrated.
- Boundary markers: The skill lacks explicit delimiters or instructions for the agent to ignore natural language instructions embedded in the provided code snippets.
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to the user to add an MCP server for documentation access via a vendor-hosted URL at docs.zenml.io.
- Source: The domain docs.zenml.io is the official documentation site for the vendor (zenml-io).
- Execution: This is a recommendation for the user to add a tool to their environment to facilitate access to official documentation.
Audit Metadata