---
title: "One command from prompt to tracked link: the CLI signs AI agents up"
description: "AI agents can now sign up for Linkbreakers from the CLI in one command, get an API token, and create links and read analytics in under 10 seconds. The CLI now covers every API operation."
canonical: "https://linkbreakers.com/help/article/one-command-from-prompt-to-tracked-link"
last-updated: 2026-10-21
---

# One command from prompt to tracked link: the CLI signs AI agents up

> AI agents can now sign up for Linkbreakers from the CLI in one command, get an API token, and create links and read analytics in under 10 seconds. The CLI now covers every API operation.

## Short answer

Ask any AI coding agent to "create a link using Linkbreakers" and it now gets there on its own. `npx -y linkbreakers-cli signup --email you@example.com` creates a workspace and saves its API token in one command, and the agent goes straight on to creating links and reading analytics. The whole thing takes under 10 seconds. You claim the workspace with a link and a code we email you. The CLI also gained a command for every API operation.

## What we shipped

### Sign up from the CLI

`linkbreakers signup --email <email>` creates a free workspace, saves an API token in the CLI config, and emails that address a link and a 6-digit code to claim it. There is no dashboard step and no key to copy. Until the workspace is claimed, every command reminds the agent of the 24-hour deadline, and once you claim it the saved token is upgraded automatically.

Already have an account? The CLI says so too: with no token configured, it offers both routes, a token from your dashboard or `signup`.

### The CLI follows the API

Every API operation is now a CLI command, generated from the same OpenAPI document as the SDKs: `linkbreakers links create`, `linkbreakers workflow-steps update`, `linkbreakers analytics-data get-top-links` and about 170 more. Flags are named after the request's fields and `--help` lists them. When the API gains an operation, the next CLI release has the command, so the CLI can no longer fall behind.

### Errors agents can act on

A failed command prints the API's error code and its message, which says what to do next. An agent that hits a limit or a paused workspace knows exactly which command fixes it.

## Frequently asked questions

### How fast is it really?

In our tests, downloading the CLI with `npx`, signing up, creating a link and reading analytics took about 5 seconds in total. Signing up alone takes about a second.

### What happens to the workspace if I do nothing?

It is deleted with everything in it 24 hours after signup. Open the emailed link and enter the code to keep it.

### Did the CLI commands change?

Yes. Commands now follow the API, so some changed shape: resources are passed by flag (`linkbreakers links get --id <id>`), list values repeat their flag (`--tags a --tags b`), and `raw` is gone because every operation has its own command. Output is JSON.

### Where do I start?

[From one prompt to a tracked link in seconds](/help/article/from-one-prompt-to-a-tracked-link-in-seconds) walks through the flow, and [How to install and use the Linkbreakers CLI](/help/article/linkbreakers-cli-setup) covers the rest.