allow disabling query model and provider

Author: zszabo-rh

README.md

@@ -369,6 +369,19 @@ customization:
   disable_query_system_prompt: true
 ```
 
+### Disable model/provider overrides from requests
+
+By default, clients may specify `model` and `provider` in `/v1/query` and `/v1/streaming_query` requests to select a model at runtime. You can disable this behavior and enforce the configured defaults by setting:
+
+```yaml
+customization:
+  disable_query_model_override: true
+```
+
+When enabled, requests that include `model` or `provider` will be rejected with HTTP 422 and the following message:
+
+"This instance does not support overriding model/provider in the query request (disable_query_model_override is set). Please remove the model and provider fields from your request."
+
 ## Safety Shields
 
 A single Llama Stack configuration file can include multiple safety shields, which are utilized in agent

docs/openapi.json

@@ -1252,6 +1252,11 @@
                         "title": "Disable Query System Prompt",
                         "default": false
                     },
+                    "disable_query_model_override": {
+                        "type": "boolean",
+                        "title": "Disable Query Model Override",
+                        "default": false
+                    },
                     "system_prompt_path": {
                         "anyOf": [
                             {