get_role_config
Resolve the provider and model for each pipeline role from environment variables, model registry, and optional plan config to verify what a plan will run on before execution.
Instructions
Show the resolved (provider, model) for every pipeline role - overlord, planner, dispatch, review, decompose - given the current env vars and model_registry.json, optionally layered with a specific plan's role_config (pass plan_name to include it). Lets you check what a plan will actually run on before executing it. Pure read; makes no changes.
"planner" here reports its own explicit configuration layer (env var / plan role_config / registry) using the same "claude" bottom-of-chain default as the other roles - it does NOT reproduce the extra "mirror dispatch's own backend when nothing else is configured" fallback that _resolve_planner_backend applies at actual dispatch time (that fallback depends on a specific story's already-resolved dispatch backend, which doesn't exist outside of a real dispatch call).
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| plan_name | No |
Output Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||