td-004-python-cli-kubernetes-workflow-separation.md

TD004: Python CLI and Kubernetes Workflow Separation

Status

Open

Scope

environment orchestration and user workflow

Summary

The Python CLI is strongly oriented around local container lifecycle and does not provide a comparable first-class orchestration path for Kubernetes environments.

Evidence

• src/vllm-sr/cli/core.py
• src/vllm-sr/cli/docker_cli.py
• docs/agent/environments.md
• deploy/operator/api/v1alpha1/semanticrouter_types.go

Why It Matters

• The Python CLI is strongly oriented around local container lifecycle and does not provide a comparable first-class orchestration path for Kubernetes environments.
• This creates an environment split where local users and Kubernetes users learn different control surfaces and config flows.
• It also makes it harder to provide one consistent product story across local dev, cluster deployment, and dashboard operations.

Desired End State

• The CLI and environment management model expose a more consistent experience across local and Kubernetes workflows.
• Environment differences are treated as deployment backends, not separate product surfaces.

Exit Criteria

• Kubernetes deployment and lifecycle management have a coherent path within the shared CLI or a clearly unified orchestration interface.
• Users do not need to mentally switch between unrelated environment management models for common operations.