Set Up NAPT in 30 Minutes: From Empty Terminal to Your First Automated Check

You do not need a lab, a server, or admin rights on the network to start automating. You need Python, an isolated environment, and twenty minutes of focus. This guide takes you from a blank terminal to a real, read-only check against a device — with the safety rails on the whole way.

1. Create an isolated environment

Always install into a virtual environment so NAPT's dependencies never collide with your system Python. One directory, fully disposable.

python3 -m venv napt-env
source napt-env/bin/activate   # Windows: napt-env\Scripts\activate
python -m pip install --upgrade pip

2. Install NAPT

The core engine pulls in the automation libraries it orchestrates.

pip install napt
napt --version

3. Provide device credentials

NAPT reads connection details from an inventory file. Keep secrets out of source control — use environment variables for passwords and reference them from the inventory.

# inventory.yaml
devices:
  - name: core-sw-1
    host: 10.0.0.11
    platform: cisco_ios
    username: ${NET_USER}
    password: ${NET_PASS}

export NET_USER="readonly"
export NET_PASS="••••••••"

4. Run your first check — safely

The most important habit you can build is --dry-run. It connects, gathers, and reports what it would do without sending a single change.

napt run check-interfaces --inventory inventory.yaml --dry-run

You should see structured output: each interface, its state, and any flagged errors.

The mental model

NAPT never lives on your devices. It runs from your workstation, opens a session per device using Netmiko or NAPALM, collects output, parses it into structured findings, and hands those findings to the rest of the platform. Removing NAPT removes nothing from the network.

Where to go next

1. Run the same check without --dry-run once you trust the output.
2. Add a second device to the inventory and watch NAPT fan out.
3. Open the Script Generator and describe a task in plain English.