agents/control-plane/internal/templates/python

# {{.ProjectName}}

A Hanzo Agents agent created with `af init`.

## Quick Start

1. **Install dependencies:**
   ```bash
   pip install -r requirements.txt
   ```

2. **Start the Hanzo Agents server** (in another terminal):
   ```bash
   af server
   ```

3. **Run the agent:**
   ```bash
   python main.py
   ```

   The agent will auto-discover an available port and register with Hanzo Agents.

## Test the Agent

### Echo Reasoner (No AI Required)
```bash
curl -X POST http://localhost:8080/v1/execute/{{.NodeID}}.demo_echo \
  -H "Content-Type: application/json" \
  -d '{"input": {"message": "Hello World"}}'
```

### Enable AI Features

1. Uncomment the `ai_config` in `main.py`
2. Set your API key: `export OPENAI_API_KEY=sk-...`
   - Or use `ANTHROPIC_API_KEY`, `GOOGLE_API_KEY`, etc. for other providers
   - LiteLLM auto-detects the provider from the model name
3. Uncomment the `analyze_sentiment` reasoner in `reasoners.py`
4. Restart the agent

```bash
curl -X POST http://localhost:8080/v1/execute/{{.NodeID}}.demo_analyze_sentiment \
  -H "Content-Type: application/json" \
  -d '{"input": {"text": "I love this product!"}}'
```

## Project Structure

- `main.py` - Agent configuration and startup
- `reasoners.py` - Reasoner definitions (AI-powered functions)
- `requirements.txt` - Python dependencies
- `.env.example` - Environment variable examples

## Learn More

- [Hanzo Agents Documentation](https://hanzo_agents.ai/docs)
- [SDK Reference](https://hanzo_agents.ai/docs/sdk)

---

Created by {{.AuthorName}} ({{.AuthorEmail}}) on {{.CreatedAt}}