Skip to content

Scope

  • Decision tables for the inference engine and the LLM gateway.
  • Fixed per-tier defaults for UI, identity, observability, orchestration, state stores and artifact registry.
  • Sizing from organization answers, a transfer manifest, and lint.
  • Deployment files such as Docker Compose or Helm. The skill decides; it does not operate. For running each tool, catalog entries point to existing per-tool skills.
  • A recommendation of which model to run.
  • Decision tables for data, CI/CD and security layers.
  • Compliance mappings.