create-lab-guide
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- External Content Ingestion: The skill clones repositories from the dockersamples organization on GitHub. This introduces remote content into the agent context for processing.
- Indirect Prompt Injection Surface: The skill parses README.md and labspace.yaml files from the cloned repositories. Ingestion point: remote repository files; Boundary markers: absent; Capability inventory: Git cloning, file system writes, and executing Docker build/lint tools; Sanitization: none specified. The risk is considered within standard operational parameters for vendor-provided content.
- Command Execution: It runs shell commands for repository cloning (git clone) and validation (docker buildx, npx prettier). These are standard tools for managing and linting documentation content.
- File System Interaction: The skill generates and writes markdown files to the content/guides/ directory, which is the intended output of the guide creation process.
Audit Metadata