zenml-quick-wins
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes various ZenML CLI commands (e.g.,
zenml status,zenml stack list,zenml secret get) and Git commands to investigate the environment and manage feature branches. These are standard operations for its intended purpose. - [EXTERNAL_DOWNLOADS]: Recommends the installation of the ZenML Docs MCP server from the official vendor domain (
https://docs.zenml.io/~gitbook/mcp). It also references the well-knowncdn.jsdelivr.netfor the Chart.js library used in HTML reports. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the user's Python codebase and ZenML stack configuration during the investigation phase. This information is passed to subagents to generate recommendations, which represents a potential injection surface if the analyzed code contains adversarial instructions.
- [DYNAMIC_EXECUTION]: The skill implements improvements by generating and applying Python code snippets and stack configuration changes based on templates provided in the
quick-wins-catalog.mdfile.
Audit Metadata