The Agent Skills Directory

Read-Only Operations: The skill is explicitly designed for discovery using gcloud asset search-all-resources and gcloud org-policies describe --effective. It includes strict instructions (Tier R) to avoid mutating commands, ensuring compliance with the principle of least privilege during environmental assessments.
Official Vendor Resources: All external references, including API endpoints (assuredworkloads.googleapis.com) and documentation links (cloud.google.com), point to official domains owned by the vendor. This ensures that the skill interacts only with authoritative and expected infrastructure.
Localized Script Execution: The included Python scripts (evaluate_control_rules.py and validate_workload_payload.py) perform data analysis on JSON telemetry files locally using standard libraries. They do not utilize dynamic code execution (like eval() or exec()) or perform unauthorized network operations.
Secure Credential Handling: The skill leverages standard Google Cloud authentication mechanisms (gcloud auth login) and Application Default Credentials. It does not require or include hardcoded secrets, API keys, or sensitive configuration files.

assuredworkloads-environment-assessor