> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.agentverse.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.agentverse.ai/_mcp/server.

# Create a Local Agent (uAgent)

> Get started building a local uAgent using the uAgents Framework

## Overview

A **local agent** (also called a uAgent) is an Agent you build and run on your own infrastructure - your local machine, a server, or the cloud. Unlike [Hosted Agents](/documentation/create-agents/hosted-agents), which run directly inside Agentverse, local agents run independently and connect to Agentverse and ASI:One once they implement the [Agent Chat Protocol (ACP)](/documentation/getting-started/enable-chat-protocol).

The **uAgents Framework** is Fetch.ai's open-source Python library for building local agents. It handles agent identity, communication, and protocol implementation out of the box.

## Prerequisites

* Python 3.8+
* PIP - the Python package manager

## Install the uAgents Framework

```bash
pip install uagents
```

Verify the installation was successful:

```bash
pip show uagents
```

## Full Framework Documentation

The uAgents Framework has comprehensive documentation covering agent creation, communication, protocols, storage, and example agents:

* [uAgents Installation Guide](https://uagents.fetch.ai/docs/getting-started/install)
* [Create your first Agent](https://uagents.fetch.ai/docs/getting-started/create)
* [ASI:One compatible Agent example](https://uagents.fetch.ai/docs/examples/asi-1)

## Next step: Connect your Agent to Agentverse

Once your local uAgent implements the Agent Chat Protocol, you can register it on Agentverse to make it discoverable and accessible through ASI:One.

Follow the [uAgents setup guide](/documentation/launch-agents/agentverse-sdk/u-agents) to connect your Agent.