Show Loops workload planes in CLI

[pstefa1707]

Summary

• Add --verbose output to truss loops push with trainer and sampler workload planes
• Add --verbose workload-plane columns to truss loops view
• Include workload-plane fields in loops view JSON output

Testing

• CI: all required PR checks are passing
• uv run ruff check truss/cli/loops_commands.py truss/tests/cli/test_loops_cli.py
• uv run pytest truss/tests/cli/test_loops_cli.py -v
• Reset local Baseten dev DB with ZERO_DOWNTIME_MIGRATIONS_RAISE_FOR_UNSAFE=False ./bin/create_dev_db inside the Baseten devcontainer.
• Confirmed Baseten's documented local training-bucket setup is AWS-backed: create_training_credentials_for_local_dev still calls the TRAINING_BUCKET_AWS_* Lambda/STS path and failed nonfatally without valid training-bucket AWS credentials (UnrecognizedClientException). There is not a self-contained local training-bucket emulator in the docs.
• Ran local Baseten API on http://localhost:8000 with in-memory local-only mocks for training-bucket credentials, checkpoint manifest S3 writes, workload-plane/MCM checkpoint secret sync, Temporal scheduling, weight mirroring, and deploy task execution. Truss maps this to REST http://api.localhost:8000.
• Seeded local Loops support for Qwen/Qwen3-8B, project p7qr9qv, ORG_ENABLE_LOOPS, and local A10G -> default workload-plane config.
• Full local CLI E2E succeeded:
  • uv run truss loops push 'Qwen/Qwen3-8B' --project-id p7qr9qv --remote baseten-localhost --verbose --non-interactive
  • Output included:
    • Run ID: gyqvxw8
    • Trainer Workload Plane: default (local, CLUSTER_LOCAL)
    • Sampler Deployment ID: gyqvxw8
    • Sampler Workload Plane: default (local, CLUSTER_LOCAL)
    • Sampler Base URL: https://model-n4q95w5d.api.baseten.co/deployment/gyqvxw8/sync
• Verified local view output:
  • uv run truss loops view --remote baseten-localhost --verbose shows the new trainer/sampler workload-plane columns.
  • uv run truss loops view --remote baseten-localhost -o json returns newline-delimited JSON objects with top-level workload_plane and nested sampler.workload_plane objects containing id, name, region, and platform.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

Paras Stefanopoulos seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}