> ## Documentation Index
> Fetch the complete documentation index at: https://docs.getbindu.com/llms.txt
> Use this file to discover all available pages before exploring further.

# 1.7 AG2 Question Answering Agent

> Simple Q&A agent built on AG2 (formerly AutoGen), wrapped with Bindu

Simple Q\&A agent that answers questions on demand.

## Code

```python theme={null}
"""AG2 (formerly AutoGen) agent wrapped with Bindu.

A simple Q&A agent using AG2's ConversableAgent, exposed as a
Bindu-compatible networked service with identity, communication,
and payment capabilities.
"""

import os

from autogen import ConversableAgent, LLMConfig
from bindu.penguin.bindufy import bindufy
from dotenv import load_dotenv

load_dotenv()

llm_config = LLMConfig(
    {
        "model": os.getenv("LLM_MODEL", "openai/gpt-4o-mini"),
        "api_key": os.environ.get("OPENROUTER_API_KEY", ""),
        "base_url": "https://openrouter.ai/api/v1",
    }
)

config = {
    "author": "ag2-community",
    "name": "ag2-assistant",
    "description": (
        "A research assistant powered by AG2 (formerly AutoGen). "
        "Ask any question and get a clear, concise answer."
    ),
    "deployment": {
        "url": "http://localhost:3773",
        "expose": True,
        "cors_origins": ["http://localhost:5173"],
    },
    
}


def handler(messages: list[dict[str, str]]):
    """Forward messages to the AG2 agent and return its response."""
    if not messages:
        return [{"role": "assistant", "content": "No input provided."}]

    user_input = messages[-1].get("content", "")
    if not user_input:
        return [{"role": "assistant", "content": "Empty message."}]

    # Fresh agents per request to avoid state leakage
    agent = ConversableAgent(
        name="assistant",
        system_message=(
            "You are a helpful research assistant. Provide "
            "clear, concise answers to questions. When you "
            "don't know something, say so."
        ),
        llm_config=llm_config,
    )
    user_proxy = ConversableAgent(
        name="user", human_input_mode="NEVER",
    )

    result = user_proxy.initiate_chat(
        agent, message=user_input, max_turns=1
    )
    if not result.chat_history:
        return [{"role": "assistant", "content": "No response."}]
    reply = result.chat_history[-1].get("content", "")
    return [{"role": "assistant", "content": reply}]


if __name__ == "__main__":
    bindufy(config, handler)
```

## How It Works

**Instructions**

* Defines agent as helpful research assistant
* Provides clear, concise answers to questions
* Admits when information is unknown
* Maintains professional, helpful tone

**Tools**

* `ConversableAgent`: Multi-agent conversation framework
* `UserProxy`: Handles message forwarding without human input
* Fresh agent instances per request prevent state leakage

**Model**

* `gpt-4o-mini`: Cost-effective model for Q\&A
* Via OpenRouter API integration
* Configurable through environment variables

## Dependencies

```bash theme={null}
uv init

uv add bindu pyautogen python-dotenv
```

## Environment Setup

Create `.env` file:

```bash theme={null}
OPENROUTER_API_KEY=your_openrouter_api_key_here
LLM_MODEL=openai/gpt-4o-mini
```

## Run

```bash theme={null}
uv run ag2-agent.py
```

**Examples:**

* "What are the main differences between AG2 and AutoGen?"
* "How do I create a multi-agent conversation with AG2?"
* "What are the key features of AG2 agents?"

## Example API Calls

<AccordionGroup>
  <Accordion title="Message Send Request">
    ```json theme={null}
    {
      "jsonrpc": "2.0",
      "method": "message/send",
      "params": {
        "message": {
          "role": "user",
          "kind": "message",
          "messageId": "9f11c870-5616-49ad-b187-d93cbb100001",
          "contextId": "9f11c870-5616-49ad-b187-d93cbb100002",
          "taskId": "9f11c870-5616-49ad-b187-d93cbb100003",
          "parts": [
            {
              "kind": "text",
              "text": "What are the main differences between AG2 and AutoGen?"
            }
          ]
        },
        "configuration": {
          "acceptedOutputModes": ["application/json"]
        }
      },
      "id": "9f11c870-5616-49ad-b187-d93cbb100003"
    }
    ```
  </Accordion>

  <Accordion title="Task get Request">
    ```json theme={null}
    {
      "jsonrpc": "2.0",
      "method": "tasks/get",
      "params": {
        "taskId": "9f11c870-5616-49ad-b187-d93cbb100003"
      },
      "id": "9f11c870-5616-49ad-b187-d93cbb100004"
    }
    ```
  </Accordion>
</AccordionGroup>

## Frontend Setup

```bash theme={null}
# Clone the Bindu repository
git clone https://github.com/GetBindu/Bindu

# Navigate to frontend directory
cd frontend

# Install dependencies
npm install

# Start frontend development server
npm run dev
```

Open [http://localhost:5173](http://localhost:5173) and try to chat with the AG2 assistant