The Agent Skills Directory

Automated Resource Discovery: The skill employs gcloud CLI commands (such as run services list and asset search-all-resources) to autonomously identify active services and workloads. This is used to streamline the monitoring setup and is a standard administrative practice for infrastructure-aware agents.
Official Documentation Access: It retrieves configuration details and metric mappings from official Google Cloud documentation domains (e.g., docs.cloud.google.com). Accessing these trusted sources ensures that the generated PromQL queries are accurate and follow current provider specifications.
Infrastructure-as-Code Generation: The primary output of the skill is Terraform (.tf) files generated via the write_to_file tool. This allows users to review and version-control their monitoring infrastructure before deployment.
Data Processing Surface: The skill ingests metadata from the environment and user-provided targets to populate templates. This is a common pattern for configuration wizards.
Ingestion points: Resource names and labels retrieved from gcloud output, as well as user-specified SLO targets and metric names.
Boundary markers: The skill uses predefined PromQL and Terraform templates as structural boundaries for data interpolation.
Capability inventory: The skill uses gcloud for environment inspection, read_url_content for retrieving metric guidelines, and write_to_file for configuration output.
Sanitization: The skill relies on structured templates to ensure that environment data is correctly formatted into the resulting observability code.

google-cloud-slo-alert-configuration