---
title: "Quickstart: agent-ready CLI"
canonical: https://tansohq.com/quickstart
description: "Tanso finds where your customers' AI agents get stuck, and fixes it. Agent Usability tests the path with the free agent-ready CLI, and Agent Monetization is the open-source billing for the payment step."
---

# Quickstart: agent-ready CLI

Check whether an AI agent can find, sign up for, and use your product. The agent-ready CLI, [@tansohq/agent-ready](https://www.npmjs.com/package/@tansohq/agent-ready) (version 0.3.0), is free. Source: [tansohq/agent-ready-cli](https://github.com/tansohq/agent-ready-cli), MIT. Needs Node 20.9 or later.

## Give this to your coding agent

    Read https://raw.githubusercontent.com/tansohq/agent-ready-cli/main/skills/agent-ready/SKILL.md and follow it to audit yourproduct.com.

## Status

- CLI: available now, free
- Hosted: available now at app.tansohq.com (sign in)

## Audit your product

    npx @tansohq/agent-ready audit yourproduct.com

Free. No key, no account. Audit reads public pages only, with GET requests, at most 36 per run. It shows seven steps: Discover, Understand, Sign up, Access, Use, Pay, Manage. It asks three questions about how agents should onboard to your product. Then it writes one fix prompt per gap, for a coding agent to apply. Audit sends nothing to Tanso.

## Verify with a real agent

    npx @tansohq/agent-ready verify

A real agent, your own Claude Code, tries the task on your product. A separate check then calls your product's API with the agent's key, with no key, and with a wrong key.

- Needs Claude Code, signed in.
- Model use is capped at $5 by default. Change it with `--max-budget-usd`. Runs in October 2026 cost $0.06 to $0.47.
- Verify creates a real account on your product.
- `npx @tansohq/agent-ready verify --check` shows the plan for free.

## Runs and claiming

Verify runs are counted on a free Tanso workspace. The first verify asks before it creates one, then saves its key on your machine.

- 6 starting runs.
- To claim the workspace, a person runs `npx @tansohq/agent-ready account --claim` to get a claim code, signs in at https://app.tansohq.com, opens Account, and enters the code.
- After claiming: 10 runs a day with the same key.
- Paid runs are planned. There is no price yet.

Details: [Access, runs and claiming](https://tansohq.com/auth.md).

## In your coding agent

Paste this prompt:

    Read https://raw.githubusercontent.com/tansohq/agent-ready-cli/main/skills/agent-ready/SKILL.md and follow it to audit yourproduct.com.

Or install it. Claude Code:

    claude plugin marketplace add tansohq/agent-ready-cli
    claude plugin install agent-ready@agent-ready

Codex, Cursor and others:

    npx skills add tansohq/agent-ready-cli

Verify starts only when you ask for it. In Claude Code, use `/agent-ready-verify`.

## CI

    npx @tansohq/agent-ready audit $URL --fail-on high --json

Exits with code 1 when the audit finds a high-severity gap.

## What Tanso receives

Audit sends nothing to Tanso. For verify, Tanso receives a signup label, that a run started, and its outcome, with your IP address. Tanso never receives your product, the task, or the agent's work.

## Hosted dashboard

1. [Sign in at app.tansohq.com](https://app.tansohq.com) with an email code.
2. Run a check. Checks are free and unlimited.
3. Save a task as a live test. A real agent runs it in your product on Tanso's model, with an independent outcome check.
4. Free: 10 hosted live tests a day per signed-in workspace. Paid runs are planned; there is no price yet.
5. To claim a CLI workspace, open the Account page and enter the code from `npx @tansohq/agent-ready account --claim`. A workspace an agent created, through the CLI or POST /v1/signup, cannot start hosted live tests until a person claims it. Until then it runs tests with the CLI: 6 starting runs, on the user's own Claude Code. After the claim, the same key gets 10 runs a day.

## More

- [README on GitHub](https://github.com/tansohq/agent-ready-cli#readme)
- [Full site context for agents](https://tansohq.com/llms-full.txt)