fix(responses): use proper text.format object for plain-text verbosity

Mypy reported a typing error because the `text.format` TypedDict expects an object
(e.g. {"type": "text"}) not a plain string "text". Previously we set
response_format = {"format": "text", "verbosity": ...} which fails type checks.

This change uses the correct format object:
  {"format": {"type": "text"}, "verbosity": <value>}

This resolves the mypy failure and keeps the `text` payload valid for the Responses API.

Author: mshsheikh

src/agents/models/openai_responses.py

@@ -299,9 +299,12 @@ async def _fetch_response(
             if response_format is not omit:
                 response_format["verbosity"] = model_settings.verbosity  # type: ignore [index]
             else:
-                # When no output_schema is present, text needs an explicit format key.
-                # Use the plain text format with verbosity to produce a valid `text` payload.
-                response_format = {"format": "text", "verbosity": model_settings.verbosity}
+                # When no output_schema is present, `text` needs an explicit format object.
+                # Use the plain text format object with verbosity to produce a valid `text` payload.
+                response_format = {
+                    "format": {"type": "text"},
+                    "verbosity": model_settings.verbosity,
+                }
 
         stream_param: Literal[True] | Omit = True if stream else omit