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

# Initialise

> Create agents.yaml configuration file

## Set API Key

<Steps>
  <Step title="Configure a provider">
    Run the interactive setup wizard to pick a provider and save your API key:

    ```bash theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
    praisonai setup
    ```

    PraisonAI supports OpenAI, Anthropic, Google/Gemini, Groq, Cohere, Ollama, OpenRouter and 100+ models via LiteLLM. Alternatively, export a key directly:

    ```bash theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
    export OPENAI_API_KEY="your-key"       # OpenAI
    export ANTHROPIC_API_KEY="your-key"   # Anthropic Claude
    export GEMINI_API_KEY="your-key"      # Google Gemini
    ```
  </Step>

  <Step title="Generate agents.yaml">
    ```bash theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
    python -m praisonai --init "your topic here"
    ```

    <Note>
      If no provider is configured, `--init` prints setup guidance and exits instead of crashing. Run `praisonai setup` (step above) to configure a provider first.
    </Note>
  </Step>
</Steps>

## Generate agents.yaml

```bash theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
python -m praisonai --init "your topic here"
```

**Example:**

```bash theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
python -m praisonai --init "research AI trends and write a summary"
```

**Output:**

```
File /path/to/agents.yaml created successfully
```

## Generated File

The command creates `agents.yaml` in the current directory:

```yaml theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
framework: praisonai
topic: research AI trends and write a summary
roles:
  researcher:
    backstory: An experienced research analyst...
    goal: Gather relevant information on the topic
    role: Research Analyst
    tasks:
      research_task:
        description: Conduct research on the topic
        expected_output: Key insights and references
    tools:
    - internet_search
```

## Next Step

Run the agents:

```bash theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
python -m praisonai
```