The Agent Skills Directory

[COMMAND_EXECUTION]: The Python scripts gws_doctor.py, workspace_audit.py, auth_setup_guide.py, and gws_recipe_runner.py use subprocess.run() to execute commands via the gws CLI. This is the intended behavior for automating Google Workspace administration tasks.
[EXTERNAL_DOWNLOADS]: The skill documentation recommends installing the gws CLI from trusted sources, specifically referencing the official @anthropic/gws npm package and the Google Workspace GitHub repository.
[INDIRECT_PROMPT_INJECTION]: The skill interfaces with external Google Workspace data, including Gmail messages, Drive files, and Calendar events. While this data ingestion creates an attack surface, the risk is mitigated by the skill's reliance on structured CLI commands and JSON parsing in output_analyzer.py rather than directly interpolating untrusted data into instructions.
[SAFE]: The skill employs secure practices for managing sensitive information, using placeholders for credentials in templates and recommending environment variables or .env files for authentication instead of hardcoded secrets.

google-workspace-cli