A smarter Playwright MCP alternative for AI agents
Playwright MCP hands your agent raw browser primitives and selectors to stitch together. Kane CLI takes a plain-English objective, drives a real Chrome browser, self-heals on UI changes, and returns a verified pass or fail. Free to install.
or read the documentation
Why teams move from Playwright MCP to Kane CLI
Playwright MCP exposes browser primitives, but stops there. The agent still plans every click, tracks page structure, recovers stale selectors, and reads raw output to judge results. That is a lot of reasoning spent on plumbing.
Kane CLI moves that work into the tool. You give it intent, not instructions. It drives a real Chrome browser, adapts when the UI shifts, asserts the result, and reports a verified pass or fail with video evidence.
You keep the escape hatch too: any flow exports to native Playwright with one command, so you are never locked out of the code underneath.
Kane CLI vs Playwright MCP
The same browser, a very different contract with your agent.
| Capability | Kane CLI (TestMu AI) | Playwright MCP |
|---|---|---|
| How the agent drives the browser | Plain-English objectives | Low-level tool calls: click, type, snapshot |
| How elements are targeted | Plain-English intent, no refs or selectors | Agent stitches accessibility-tree refs per snapshot |
| Resilience when the UI changes | Adapts and retries up to 50 steps | |
| Verified pass or fail | Built-in assertions, returns pass or fail | Agent must interpret raw output |
| Runs without an MCP client | Works from any terminal or CI | Designed to be driven by an MCP client |
| Cross-browser and OS on the cloud | Run across browsers and OS on the TestMu AI grid | |
| Replayable, cached tests | test.md replays with no repeat LLM cost | |
| Automated bug discovery | Crawls and surfaces bugs while running | |
| Evidence and reporting | Video, step trace, shareable links, dashboard | Console and log output |
| Native Playwright export | One command to export Playwright code | Calls Playwright directly under the hood |
What you get with Kane CLI
Everything Playwright MCP leaves to you, handled by the tool.
Natural language, not tool calls
Describe the journey once. Kane CLI plans the steps, drives Chrome, and verifies the outcome. No selectors, glue code, or brittle scripts.
Resilient by default
When the frontend changes, Kane CLI adapts on its own and pushes through up to 50 steps until the full journey is verified.
Verified pass or fail
Built-in assertions check each step, not just the final screen, and flag the exact point of failure so your agent knows what to do next.
Agent-native output
The --agent flag emits structured NDJSON your AI coding agent parses directly, with screenshots to read on failure.
Run anywhere
From your terminal, in CI, or inside an agent. Local runs are free, and one flag scales to cross-browser cloud runs.
Own the Playwright code
Hacker mode enabled. Export any validated flow to native Playwright, then modify and own it. The convenience does not lock you in.
Build up confidence locally
Start in your terminal
Validate on the cloud
Release with confidence
Built for agents and humans, on one engine
Kane CLI and KaneAI share the same automation engine and dashboard.
Same engine, two front doors
Kane CLI is the terminal-native companion to KaneAI. Humans run flows from the shell, agents through --agent, and both sync to the same dashboard.
Built for the agent loop
Generate code, then verify it in a real browser before opening a PR. Kane CLI closes the gap between what an agent writes and what actually works.
Evidence you can share
Every run produces a persistent video, a step trace, and a replay link to drop into a PR or bug report.
Replace Playwright MCP in three steps
Install
Run npm install -g @testmuai/kane-cli, then sign in with your TestMu AI account. No MCP server to configure per client.
Point your agent
Tell Claude Code, Cursor, Codex, or Gemini to use Kane CLI for browser work. It reads the agent guide and takes over.
Run and verify
Describe the flow in plain English. Kane CLI drives a real browser, verifies each step, and returns a pass or fail with evidence.
Get Started With Kane CLI
🎉 Launch offer: Bonus credits for the first 3 months on paid plans
Choose the right plan for you
Local test authoring via CLI
Auto-heal & vision
View test cases on UI
Test Manager
Free
$0
/month
200 Credits
Resets in every
30 days
Starter
$19
/month
2000 Credits
Launch: 4,000 Credits (+100%)
Bonus for first 3 months
Pro
$99
/month
10,000 Credits
Launch: 15,000 Credits (+50%)
Bonus for first 3 months
Enterprise
Get access to solutions built on Enterprise-Grade Security, Privacy, and Compliances.
Need more credits?
Got a bigger use case in mind?
Let’s talk
Choose the right plan for you
Free
$0
/month
200 Credits
Resets in every
30 days
Starter
$19
/month
2000 Credits
Launch: 4,000 Credits (+100%)
Bonus for first 3 months
Pro
$99
/month
10,000 Credits
Launch: 15,000 Credits (+50%)
Bonus for first 3 months
Enterprise
Get access to solutions built on Enterprise-Grade Security, Privacy, and Compliances.
Need more credits?
Got a bigger use case in mind?
Let’s talk
Get the technical rundown
Documentation
Everything you need to install, configure, and run Kane CLI in under 2 minutes.
Frequently asked questions
Kane CLI by TestMu AI is the leading Playwright MCP alternative for AI agents and developers. Instead of low-level browser tool calls, it takes a plain-English objective, drives a real Chrome browser, self-heals on UI changes, verifies every step, and returns a clean pass or fail. It runs from any terminal, in CI, or from an AI coding agent like Claude Code, Cursor, or Codex.
Playwright MCP is a Model Context Protocol server that gives an AI client browser primitives (navigate, click, type, snapshot) backed by Playwright. The agent must stitch those primitives together itself and read raw output to decide what happened. It is powerful, but resilience, assertions, and reporting are left to you.
Three big differences. First, you describe intent in natural language instead of issuing tool calls, so there are no selectors to maintain. Second, Kane CLI is resilient by default: it adapts and retries up to 50 steps when a page changes, rather than failing on a stale selector. Third, it returns a verified pass or fail with built-in assertions, video evidence, and a shareable replay link, so an agent never has to interpret raw logs.
Yes. Every run supports an --agent flag that switches output to structured NDJSON, so any AI coding agent can parse results, read a screenshot on failure, and decide what to do next. Point your agent at the published agent guide and it will install, authenticate, and drive Kane CLI on its own.
Yes. Any validated flow exports to native Playwright with one command. You get natural language during authoring and full ownership of the generated Playwright script when you want to take it further.
Yes. Authenticate with your TestMu AI credentials, pass --headless and --timeout on each run, and gate your pipeline on the exit code: 0 on pass, 1 on fail, 2 on setup or auth errors, and 3 on timeout.
The CLI is free to install and use. Local runs are free; cloud runs on the TestMu AI grid are billed against your TestMu AI plan. Start on the free tier and run end to end without a credit card.
Teach your agent the right skills
Point your AI coding agent at the Kane CLI guide and it will install, authenticate, and run verified browser flows on its own.
Point your agent to: testmuai.com/kane-cli/agents.md