 |
This terminal-native browser validation tool launches today, natively supporting Claude Code, Codex CLI, Cursor, and Gemini CLI, and is free to start using.
SAN FRANCISCO and NOIDA, IndiaApril 30, 2026 /PRNewswire/ — TestMu AI (formerly LambdaTest), the world’s first full-stack intelligent agent quality engineering platform, today announced the launch of Kane CLI, a new browser automation tool that runs directly in the terminal. Kane CLI is the first tool designed for both human developers and AI coding agents, bridging the gap between code generation and validated browser execution.
AI coding agents have revolutionized how software is written, as AI agents push code at speeds far exceeding any QA team’s ability to manually click through processes for validation. Features go live with a prompt. Bugs are fixed in seconds. But the development loop was never truly closed: no agent could open a browser and verify that what it built actually works. That step still required a human. Kane CLI was built to close that loop.
Teams build features, find bugs, push code, and run agents. Kane CLI is the validation layer for all of this. Developers and test engineers streamline processes, attaching full step-by-step links and screenshots before submitting a pull request (PR), delivering a pass or fail verdict. Designers and project managers can verify fixes and broken flows without filing tickets or waiting on developers, then drop shareable evidence links directly into Slack or Jira.
AI agents like Claude Code, Codex CLI, Cursor, and Gemini CLI build the features, while Kane CLI is the tool that fills the gap, testing whether those features actually work in a local Chrome browser.
Key Features of Kane CLI Include:
- Intent-based browser control: Operates purely on intent, without requiring selectors or underlying code.
- Resilient execution: Kane CLI doesn’t quit midway. It adapts and pushes forward, executing up to 50 steps per flow until the entire journey is validated. Other tools break at the first change. Kane CLI completes the run.
- Playwright export: Converts plain-English test flows into native Playwright test code with a single command.
- Automatic vulnerability discovery: Actively monitors and detects unexpected behavior during test flow execution.
- Vision-based dynamic waiting: Detects loading animations and dynamic effects on screen before performing actions. Not network-based. Handles Canvas, Shadow DOM, and element frameworks that can’t be parsed through conventional means.
- Handles OTP and CAPTCHA: When automation encounters an OTP screen or CAPTCHA, it doesn’t silently fail. Kane CLI pauses, requests human intervention for that step, then continues. For AI agents, this is human-in-the-loop without stopping the entire workflow.
- Bidirectional script migration: Converts existing Playwright or Selenium scripts to Kane CLI. Also converts Kane CLI tests back to Playwright. No need to rewrite from scratch.
- Built-in test manager sync: Every locally created test case is also saved remotely. Shareable evidence is automatically attached.
- CI/CD ready: Runs headlessly in GitHub Actions, GitLab CI, Jenkins, and Bitbucket Pipelines. Standard exit codes integrate directly into pipeline control flows without any custom scripts.
- Contextual authoring: Provide Kane CLI with application context, and it can write test cases in parallel across multiple browser sessions from a single prompt.
Three Operating Modes
Kane CLI offers three modes, allowing both humans and agents to use it the same way from the same terminal.
- Interactive TUI: No parameters needed. Opens a full terminal UI for exploring, iterating, and chaining tests in a live browser session.
- Headless CLI: Add the –headless flag for scriptable, display-free runs. Built for shell scripts and CI pipelines.
- Agent mode: Add the –agent –headless flags. Outputs structured NDJSON data that Claude Code, Codex CLI, and Gemini CLI can natively read to decide next steps.
Asad Khan, CEO and Co-founder of TestMu AI, said: “For years, the bottleneck in software development was writing code. Vibecoding eliminated that bottleneck. Teams are shipping software faster and in greater volume than ever before in the industry’s history. But this has exposed a new bottleneck that most teams haven’t yet named: trust. Every feature that goes live via a prompt is a feature no one has actually validated. At agentic speed, ‘a human will click through it later’ isn’t a plan—it’s a liability, and one that grows with every cycle of AI speed. It’s a growing pile of unverified work. That’s why we built Kane CLI. One terminal command, one real browser, one pass or fail result. Software has always trusted the person who wrote it. For the first time, software now has to trust a machine. Kane CLI is how trust scales in the age of agents.”
As a launch promotion, TestMu AI is offering teams that activate a paid plan during the promotional period bonus credits for the first three months. This offer is designed to give engineering and quality teams full access to Kane CLI’s cloud capabilities.
Kane CLI is now available and free to start using. Install via npm or brew, log in, and run your first flow.
Install Kane CLI via npm: npm install -g @testmuai/kane-cli
Install Kane CLI via Homebrew: brew install LambdaTest/kane/kane-cli
Building with AI agents? Visit testmuai.com/kane-cli/agents.md
For a quick start, check the documentation: testmuai.com/support/docs/kane-cli-introduction/; for more information, visit: testmuai.com/kane-cli