# Installation

Kitaru requires **Python 3.11 or newer**.

You can verify your interpreter with:

```bash
python --version
```

{% tabs %}
{% tab title="uv (recommended)" %}

```bash
uv add kitaru
```

{% endtab %}

{% tab title="pip" %}

```bash
pip install kitaru
```

{% endtab %}
{% endtabs %}

This gives you the full SDK, CLI, and everything you need to run flows locally.

## Optional extras

| Extra         | What it adds                                                  |
| ------------- | ------------------------------------------------------------- |
| `local`       | Local server and UI for browsing executions in a local web UI |
| `mcp`         | MCP server for querying executions from AI assistants         |
| `pydantic-ai` | PydanticAI adapter for wrapping agents in checkpoints         |

```bash
uv add "kitaru[mcp,pydantic-ai,local]"
# or: pip install "kitaru[mcp,pydantic-ai,local]"
```

If you use Claude Code or another MCP-capable assistant, install `kitaru[mcp]` so your assistant can query executions, inspect logs and artifacts, provide input to waiting runs, and start replays through structured tool calls. See [MCP Server](/kitaru/agent-native/mcp-server.md) for setup.

## Verify Installation

{% tabs %}
{% tab title="uv project" %}

```bash
uv run kitaru --version
uv run kitaru --help
```

{% endtab %}

{% tab title="pip environment" %}

```bash
kitaru --version
kitaru --help
```

{% endtab %}
{% endtabs %}

## Local UI

If you installed the `local` extra, you can start a local server with a web UI for browsing executions:

```bash
kitaru login
```

This launches the Kitaru server on your machine and opens the UI in your browser. You can inspect flows, checkpoints, logs, and artifacts from any execution you have run locally.

## Next Steps

Head to the [Quickstart](/kitaru/getting-started/quickstart.md) to explore what's available, see [Execution management](/kitaru/guides/execution-management.md) for lifecycle operations, or open [MCP Server](/kitaru/agent-native/mcp-server.md) for assistant-native querying.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.zenml.io/kitaru/getting-started/installation.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.