Skip to main content
AgentNet CLI is a Python package that installs the agentnet command on your system. It requires Python 3.11 or later and works on macOS, Linux, and Windows.

Install methods

pip

Installing with pip adds agentnet to your PATH permanently. This is the recommended method for most users.
If you use a virtual environment, activate it first. To install system-wide on macOS or Linux, you may need to prefix with sudo or use pip install --user agentnet-cli to install into your user directory.

uvx

uvx runs AgentNet CLI in a temporary, isolated environment without installing it persistently. This is useful if you want to try the CLI without modifying your Python environment, or if you use uv as your package manager.
When running with uvx, replace agentnet in any command with uvx agentnet. For example: uvx agentnet setup or uvx agentnet detect. You do not need to run pip install first.

From source

If you want to run the latest unreleased code or contribute to the project:
After syncing, run commands with uv run agentnet:

Verify the installation

After installing, confirm the CLI is available and check the version:
You should see output like:
If the command is not found, check that your Python scripts directory is in your PATH. With pip install --user, the scripts directory is typically ~/.local/bin on Linux/macOS or %APPDATA%\Python\Scripts on Windows.

Python version requirement

AgentNet CLI requires Python 3.11 or later. It is tested against Python 3.11, 3.12, and 3.13. To check your Python version:
Python 3.10 and earlier are not supported. If your system Python is older, use a tool like pyenv or install a newer Python version before installing AgentNet CLI.

Upgrading

To upgrade to the latest version:
Or let the CLI upgrade itself:
agentnet update checks PyPI for a newer version, upgrades if one is available, and refreshes any stale agent connection configs.

Next steps

Once installed, run agentnet setup to sign in and connect your agents:
See the quickstart guide for a step-by-step walkthrough of the setup flow.