HITL Overview

"""AG2 agent with weather and sales tools for CopilotKit showcase.Uses AG2's ConversableAgent with AGUIStream to exposethe agent via the AG-UI protocol."""from __future__ import annotationsimport jsonimport loggingfrom typing import Annotated, Anyimport openaifrom autogen import ConversableAgent, LLMConfigfrom autogen.ag_ui import AGUIStreamfrom dotenv import load_dotenvfrom pydantic import ValidationErrorload_dotenv()# Import shared tool implementationsfrom tools import (    get_weather_impl,    query_data_impl,    manage_sales_todos_impl,    get_sales_todos_impl,    schedule_meeting_impl,    search_flights_impl,    build_a2ui_operations_from_tool_call,    RENDER_A2UI_TOOL_SCHEMA,)from tools.types import Flightfrom ._header_forwarding import get_forwarded_headersfrom ._request_context import get_latest_user_messagelogger = logging.getLogger(__name__)# Module-level async client: re-used across requests (httpx connection pool is# thread-safe). Using AsyncOpenAI inside an `async def` avoids blocking the# ASGI event loop on the secondary LLM call._async_openai_client = openai.AsyncOpenAI()# =====# Tools# =====async def get_weather(    location: Annotated[str, "City name to get weather for"],) -> str:    """Get current weather for a location."""    result = get_weather_impl(location)    # Return a JSON string (not a dict): autogen serializes dict returns with    # str(), producing a Python repr (single quotes) that the frontend's    # parseJsonResult/JSON.parse cannot parse — the weather card then renders    # "--" placeholders. Same pattern as search_flights below.    return json.dumps(        {            "city": result["city"],            "temperature": result["temperature"],            "feels_like": result["feels_like"],            "humidity": result["humidity"],            "wind_speed": result["wind_speed"],            "conditions": result["conditions"],        }    )async def query_data(    query: Annotated[str, "Natural language query for financial data"],) -> str:    """Query financial database for chart data."""    # Return a JSON string (not a list): autogen serializes non-str returns    # with str(), producing a Python repr (single quotes) that the frontend's    # parseJsonResult/JSON.parse cannot parse. Same pattern as get_weather.    return json.dumps(query_data_impl(query))async def manage_sales_todos(    todos: Annotated[list, "Complete list of sales todos"],) -> str:    """Manage the sales pipeline."""    # See contract comment on query_data above — return JSON, not dict.    # SalesTodo is a Pydantic model; coerce via model_dump for serialisability.    result = [t.model_dump() for t in manage_sales_todos_impl(todos)]    return json.dumps({"todos": result})async def get_sales_todos() -> str:    """Get the current sales pipeline."""    # See contract comment on query_data above — return JSON, not list.    # SalesTodo is a Pydantic model; coerce via model_dump for serialisability.    return json.dumps([t.model_dump() for t in get_sales_todos_impl(None)])async def schedule_meeting(    reason: Annotated[str, "Reason for the meeting"],) -> str:    """Schedule a meeting with user approval."""    # See contract comment on query_data above — return JSON, not dict.    return json.dumps(schedule_meeting_impl(reason))async def search_flights(    flights: Annotated[        list[dict[str, Any]], "List of flight objects to display as rich A2UI cards"    ],) -> str:    """Search for flights and display the results as rich cards. Return exactly 2 flights.    Each flight must have: airline, airlineLogo, flightNumber, origin, destination,    date (short readable format like "Tue, Mar 18" -- use near-future dates),    departureTime, arrivalTime, duration (e.g. "4h 25m"),    status (e.g. "On Time" or "Delayed"),    statusColor (hex color for status dot),    price (e.g. "$289"), and currency (e.g. "USD").    For airlineLogo use Google favicon API:    https://www.google.com/s2/favicons?domain={airline_domain}&sz=128    """    try:        typed_flights: list[Flight] = [Flight(**f) for f in flights]    except ValidationError as exc:        logger.warning(            "search_flights: invalid flight shape type=%s err=%s",            type(exc).__name__,            exc,            exc_info=True,        )        return json.dumps({"error": f"invalid flight shape: {exc}"})    result = search_flights_impl(typed_flights)    return json.dumps(result)async def generate_a2ui(    context: Annotated[str, "Conversation context to generate UI for"],) -> str:    """Generate dynamic A2UI components based on the conversation.    A secondary LLM designs the UI schema and data. The result is    returned as an a2ui_operations container for the middleware to detect.    """    # A13: AsyncOpenAI inside async def (was sync openai.OpenAI which blocks    # the ASGI event loop). Forward x-* headers via extra_headers in addition    # to the global httpx hook so aimock context routing is explicit at the    # call site.    #    # R2-A1 / A4: thread the latest user prompt from the inbound    # RunAgentInput.messages payload (captured into a per-request ContextVar    # by RequestUserMessageMiddleware — see agents/_request_context.py) into    # the inner LLM call so each pill's request body is byte-distinct.    # Without this, every pill landing on the omnibus agent (agentic-chat /    # tool-rendering / chat-customization-css / hitl) produces an IDENTICAL    # inner-LLM body and the aimock fixture cannot disambiguate. Falls back    # to the original hardcoded prompt when the middleware captured nothing    # (parse failure already logged at WARNING).    user_prompt = get_latest_user_message() or (        "Generate a dynamic A2UI dashboard based on the conversation."    )    forwarded = get_forwarded_headers()    try:        response = await _async_openai_client.chat.completions.create(            model="gpt-5-mini",            messages=[                {                    "role": "system",                    "content": context or "Generate a useful dashboard UI.",                },                {                    "role": "user",                    "content": user_prompt,                },            ],            tools=[                {                    "type": "function",                    "function": RENDER_A2UI_TOOL_SCHEMA,                }            ],            tool_choice={"type": "function", "function": {"name": "render_a2ui"}},            extra_headers=forwarded or None,        )    except Exception as exc:        logger.error(            "generate_a2ui: inner LLM call failed type=%s err=%s",            type(exc).__name__,            exc,            exc_info=True,        )        return json.dumps({"error": f"inner LLM call failed: {type(exc).__name__}"})    if not response.choices:        logger.warning("generate_a2ui: LLM returned no choices")        return json.dumps({"error": "LLM returned no choices"})    choice = response.choices[0]    if not choice.message.tool_calls:        logger.warning("generate_a2ui: secondary LLM produced no render_a2ui tool call")        return json.dumps({"error": "LLM did not call render_a2ui"})    try:        args = json.loads(choice.message.tool_calls[0].function.arguments)        result = build_a2ui_operations_from_tool_call(args)        return json.dumps(result)    except (json.JSONDecodeError, KeyError, TypeError, ValueError) as exc:        logger.error(            "generate_a2ui: failed to parse render_a2ui args type=%s err=%s",            type(exc).__name__,            exc,            exc_info=True,        )        return json.dumps(            {"error": f"failed to parse render_a2ui args: {type(exc).__name__}"}        )# =====# Agent# =====agent = ConversableAgent(    name="assistant",    system_message=(        "You are a helpful sales assistant. You can look up current weather "        "for any city using the get_weather tool, query financial data with "        "query_data, manage the sales pipeline with manage_sales_todos and "        "get_sales_todos, schedule meetings with schedule_meeting, search "        "flights and display rich A2UI cards with search_flights, and "        "generate dynamic A2UI dashboards with generate_a2ui. "        "When asked about the weather, always use the tool rather than guessing. "        "Be concise and friendly in your responses."    ),    llm_config=LLMConfig({"model": "gpt-5-mini", "stream": True}),    human_input_mode="NEVER",    # Guard against infinite tool-call loops: AG2's ConversableAgent with    # human_input_mode="NEVER" will keep executing tool calls indefinitely    # if the LLM keeps requesting them.  Without this limit the agent floods    # Railway's log stream (500 logs/sec rate-limit), becomes unresponsive    # to health probes, and gets killed by the watchdog.    max_consecutive_auto_reply=15,    functions=[        get_weather,        query_data,        manage_sales_todos,        get_sales_todos,        schedule_meeting,        search_flights,        generate_a2ui,    ],)# AG-UI stream wrapperstream = AGUIStream(agent)