[Bug]: Gpt-5 Default sample failed.

[passacagllia]

Language

Javascript/Typescript, Python

Version

latest

Description

This is my code:

import os
import sys
import json
import traceback
from dataclasses import asdict

from botbuilder.core import MemoryStorage, TurnContext
from teams import Application, ApplicationOptions, TeamsAdapter
from teams.ai import AIOptions
from teams.ai.models import AzureOpenAIModelOptions, OpenAIModel
from teams.ai.planners import ActionPlanner, ActionPlannerOptions
from teams.ai.prompts import PromptManager, PromptManagerOptions
from teams.state import TurnState
from teams.feedback_loop_data import FeedbackLoopData

from config import Config

config = Config()

# Create AI components
model: OpenAIModel

model = OpenAIModel(
    AzureOpenAIModelOptions(
        api_key=config.AZURE_OPENAI_API_KEY,
        default_model="gpt-5",
        endpoint=config.AZURE_OPENAI_ENDPOINT,
        api_version="2025-04-01-preview",
    )
)

prompts = PromptManager(PromptManagerOptions(prompts_folder=f"{os.getcwd()}/prompts"))

planner = ActionPlanner(
    ActionPlannerOptions(
        model=model,
        prompts=prompts,
        default_prompt="chat",
        enable_feedback_loop=True,
    )
)

# Define storage and application
storage = MemoryStorage()
bot_app = Application[TurnState](
    ApplicationOptions(
        bot_app_id=config.APP_ID,
        storage=storage,
        adapter=TeamsAdapter(config),
        ai=AIOptions(planner=planner, enable_feedback_loop=True),
    )
)

@bot_app.error
async def on_error(context: TurnContext, error: Exception):
    # This check writes out errors to console log .vs. app insights.
    # NOTE: In production environment, you should consider logging this to Azure
    #       application insights.
    print(f"\n [on_turn_error] unhandled error: {error}", file=sys.stderr)
    traceback.print_exc()

    # Send a message to the user
    await context.send_activity("The agent encountered an error or bug.")

@bot_app.feedback_loop()
async def feedback_loop(_context: TurnContext, _state: TurnState, feedback_loop_data: FeedbackLoopData):
    # Add custom feedback process logic here.
    print(f"Your feedback is:\n{json.dumps(asdict(feedback_loop_data), indent=4)}")

It works ok with gpt-4.1. How to fix it? Thank you.

Reproduction Steps

1. Use Create Template from 365 Tool Kit from VSC. 
2. Test with gpt 4.1 ok 
3. Test with gpt 5 failed.

[askpatrickw]

Is this not a dupe of #2364, isn't GPT-5 Responses API only?