fix: normalize remaining reasoning effort orderings and add missing 'minimal'

Follow-up to cherry-picked PR #6698. Fixes spots the original PR missed:
- hermes_constants.py: VALID_REASONING_EFFORTS tuple ordering
- gateway/run.py: _load_reasoning_config docstring + validation tuple
- configuration.md and batch-processing.md: docs ordering
- hermes-agent skill: /reasoning usage hint was missing 'minimal'

website/docs/user-guide/configuration.md

@@ -747,7 +747,7 @@ Control how much "thinking" the model does before responding:
 
 ```yaml
 agent:
-  reasoning_effort: ""   # empty = medium (default). Options: xhigh (max), high, medium, low, minimal, none
+  reasoning_effort: ""   # empty = medium (default). Options: none, minimal, low, medium, high, xhigh (max)
 ```
 
 When unset (default), reasoning effort defaults to "medium" — a balanced level that works well for most tasks. Setting a value overrides it — higher reasoning effort gives better results on complex tasks at the cost of more tokens and latency.

website/docs/user-guide/features/batch-processing.md

@@ -79,7 +79,7 @@ Entries can optionally include:
 
 | Parameter | Description |
 |-----------|-------------|
-| `--reasoning_effort` | Effort level: `xhigh`, `high`, `medium`, `low`, `minimal`, `none` |
+| `--reasoning_effort` | Effort level: `none`, `minimal`, `low`, `medium`, `high`, `xhigh` |
 | `--reasoning_disabled` | Completely disable reasoning/thinking tokens |
 
 ### Advanced Options