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

# Quickstart

> Start monitoring your agents! 

## Get started in three steps

Get your documentation site running locally and make your first customization.

<Step title="Install nanomon">
  ```bash theme={null}
  pip install nanomon
  ```
</Step>

<Step title="Set environment variables" stepNumber={2}>
  Obtain project credentials in the project settings and set the environment variables:

  ```bash theme={null}
  NANOMON_CLIENT_ID=<PROJECT_CLIENT_ID>
  NANOMON_SECRET_ID=<PROJECT_CLIENT_ID>
  ```
</Step>

<Step title="Decorate your tools" stepNumber={3}>
  Decorate the tools you want to monitor.

  ```python theme={null}
  @nanomon
  def your_tool(<inputs>) -> <output>:
  	<your code>
  ```
</Step>

<Step title="Integrate with your flow" stepNumber={4}>
  Basic setup: Set the nanomon integration right before invoking your graph.

  ```python theme={null}
  with nanomon.run(
  	tags=["task_estimation_flow"], # any tags you want
  	metadata={field: value}
  ):
  graph.ainvoke(initial_state)
  ```
</Step>

<CardGroup cols={1}>
  <Card title="Self-Hosted or Cloud-Hosted" icon="block-quote" color="#e300ff" horizontal href="/essentials/markdown">
    You decide how you want to start using nanomon.
  </Card>
</CardGroup>