Kane CLI By TestMu AI

Plain-English Browser Testing

Instead of writing and maintaining brittle test scripts, you describe what should happen, such as “log in as an admin, open the billing page, and verify the plan shows Enterprise,” and Kane CLI drives a real Chrome browser to execute it, verify each step, and return a deterministic pass or fail.

• No selectors
• No custom domain-specific language
• No framework boilerplate

Just intent, expressed in natural language, run against a real browser.

Built for Humans and AI Agents

What sets Kane CLI apart is its dual identity: it is built for humans and AI agents at the same time.

Developers and QA Engineers

Developers and QA engineers use it to author end-to-end tests in plain English rather than maintaining Playwright or Selenium suites.

AI Coding Agents

AI coding agents, including Claude Code, Cursor, Codex CLI, Gemini CLI, and custom agents, invoke it to verify that the code they generate actually works in a real browser.

Platform and DevOps Teams

And platform or DevOps teams run it as a browser-based check inside CI pipelines without a test codebase to maintain.

One tool serves all three, because the interface is the same for each: describe the objective, get back a trustworthy result.

Verification at the Heart of the Product

This focus on verification is the heart of the product.

• A general browser agent can perform tasks, but it has no concept of pass or fail.
• Traditional frameworks can assert outcomes, but they shatter every time the UI shifts and demand constant upkeep.

Kane CLI bridges the two.

Every run ends in a binary, repeatable result backed by a real browser session you can replay. The same objective and context produce the same outcome, not a one-shot guess.

That reliability is what makes it a testing tool rather than just an automation novelty.

Resilient and Portable Test Execution

Several capabilities make those results resilient and portable.

Autoheal

Autoheal absorbs cosmetic UI changes, such as renamed buttons or shifted layouts, that would normally break a scripted test, pushing through to complete the full journey.

Vision-based dynamic waiting

Vision-based dynamic waiting lets Kane CLI detect what is actually rendered on screen, so it handles canvas elements, shadow DOM, and JavaScript-heavy frameworks that defeat selector-based tools.

Secure, parameterized flows

Secure, parameterized flows let you inject secrets and variables to build dynamic, reusable tests.

Custom profiles and stateful sessions

Custom profiles and stateful sessions allow testing behind authenticated logins and role-specific views.

Built-in Ask tool

And when a flow hits a CAPTCHA or OTP, the built-in Ask tool pauses, brings a human into the loop to clear that single step, and resumes, keeping an agent’s workflow moving without defeating bot protection.

Native Playwright Export

When a flow stabilizes, Kane CLI does not lock you in.

Native Playwright export, what the team calls “hacker mode,” converts any run into editable Playwright code you can drop into a maintained suite and modify freely.

This makes Kane CLI a complement to your existing stack rather than a replacement: start fast in natural language, graduate to code when you need it.

Persistent Test Evidence

Every run also produces a persistent, shareable evidence link with:

• Full replay
• Video
• Step-by-step trace
• Screenshots
• Network and console logs

This is proof you can paste straight into a Slack thread, a Jira ticket, or a pull request comment.

Three Execution Modes

Kane CLI runs in three modes to fit any context.

1. Interactive TUI: An interactive TUI lets you watch exactly what the tool does while debugging in visible mode.
2. Headless CLI: A headless CLI runs the identical flow in CI without changing a single line, using standard exit codes that integrate cleanly with GitHub Actions, GitLab CI, Jenkins, and Bitbucket Pipelines.
3. Agent mode: And agent mode exposes structured output so AI agents can drive and parse runs programmatically.

Multi-Environment Testing and Test.md

Multi-environment testing is a single command.

Point the same flow at staging or production by changing the URL, with no duplicate scripts.

There is also Test.md, an agent-native test format that captures any session as replayable markdown with imports, variables, and replay built in.

Kane CLI and KaneAI

Kane CLI is the terminal-native companion to KaneAI, TestMu AI’s web-based test authoring and management platform.

They share the same automation engine, and runs triggered from the CLI upload to the KaneAI dashboard, so you get replay, logs, and full test-case management alongside your web-authored tests.

Getting Started

Getting started takes minutes.

1. Install with npm install -g @testmuai/kane-cli
2. Authenticate against your TestMu AI account
3. Run your first flow

Pricing

• The CLI is free to install and use, with a free tier of 200 credits per month.
• Paid plans start at 19 dollars per month (Starter) and 99 dollars per month (Pro), with Enterprise plans offering enhanced security, privacy, and compliance.
• Local runs are free, and cloud runs on the TestMu AI grid bill against your plan.