TestMu AI, the leading name behind full-stack Agentic Quality Engineering, recently launched Kane CLI – an exciting browser automation framework working directly on your terminal. Interestingly, Kane is the very first platform to bridge the chasm between human software developers and their machine counterparts in AI coding assistants.
AI coding assistants have completely changed how software applications are developed in modern times, churning out codes faster than conventional manual testing approaches can keep up with. While features are now deployed via prompts and bugs are rectified in seconds, a critical gap has remained: the inability of agents to autonomously verify their output within a live browser environment. Kane CLI is designed to close this loop by serving as the essential verification layer.
The tool integrates seamlessly with industry-leading AI agents including Claude Code, Codex CLI, Cursor, and Gemini CLI providing the “missing link” that confirms whether AI-generated features function correctly in a local Chrome environment.
Redefining Browser Automation for the AI Era
Kane CLI empowers development and QA teams to describe workflows and receive immediate pass/fail results, complete with step-by-step tracing and visual evidence. This streamlined approach allows project managers and designers to verify fixes and identify regressions without developer intervention, sharing evidence directly to collaborative platforms like Slack and Jira.
Some key aspects of Kane CLI are:
- Intent Based Navigation: Runs completely based on user intent without needing any complex selectors or any underlying code.
- Resilient Workflow Execution: Whereas conventional tools get blocked on encountering the very first UI change, Kane CLI is able to adapt itself dynamically with up to 50 steps per workflow execution.
- Two-way Script Migration: Allows hassle-free migration of Kane CLI script to existing Playwright or Selenium scripts.
- Human and AI Interaction: In case of OTPs or CAPTCHAs, the tool automatically stops the execution process for human intervention.
- Vision Based Detection: Performs visual detection of loaders and animation instead of being dependent on network signals, thus compatible with Canvas, Shadow DOM, and iframes.
- Prepared for CI/CD: Provides out-of-the-box integration with GitHub Actions, GitLab CI, Jenkins, and Bitbucket while working in headless mode with exit codes.
Also Read: Cequence Security Bridges the AI Governance Gap with Launch of New “Agent Personas” for AI Gateway
A Unified Experience for Humans and Machines
The tool offers three distinct operational modes to suit any workflow:
- Interactive TUI: A full terminal UI for real-time iteration and exploration.
- Headless CLI: Optimized for shell scripts and automated CI pipelines.
- Agent Mode: Generates structured NDJSON specifically for AI agents to interpret and act upon autonomously.
Asad Khan, CEO and co-founder of TestMu AI, commented, “For years, the bottleneck in software development was writing code. Intuition programming eliminated that. Teams are releasing more software, and faster, than ever before in our industry. But this has brought to light a new bottleneck that most teams haven’t yet identified: trust. Every feature launched from a command line is a feature that no one has actually verified. At the speed of agents, ‘a human will review it later’ isn’t a plan; it’s a risk that’s compounded at the speed of AI. It’s a growing stack of unverified work. That’s why we created Kane CLI. One terminal command, one actual browser, approved or rejected. Software has always trusted the people who wrote it. Now, for the first time, it has to trust machines. Kane CLI is how trust adapts to the age of agents.”
Availability and Launch Incentives
Kane CLI is available immediately and is free to start. To support the engineering community during the launch phase, TestMu AI is offering exclusive bonus credits for the first three months to teams activating paid plans.
