Modern software delivery has fundamentally changed. What once followed a slow, linear release cycle: design → development → testing → release—has now evolved into a continuous, fast-moving pipeline powered by Agile, DevOps, and CI/CD practices. In this environment, traditional test automation is no longer enough.
Chatbots have changed the way users interact with software. Instead of clicking through rigid workflows, users now type questions, give instructions, ask for summaries, request charts, upload files, and expect intelligent responses. This shift has created a new testing problem that traditional automation tools were never designed to solve.
For years, teams have relied on tools like Selenium and Playwright to test web applications. These tools work well when the application behaves in a mostly deterministic way. A button click should open a modal. A form submission should show a success message. A table should contain a certain row. In these cases, testers can depend on locators, assertions, and exact text matching.
But chatbot testing is different.
Every engineering team knows the tension. You ship fast, but testing can't keep up. Regression suites grow exponentially. Manual QA becomes a bottleneck. And somewhere between the PR and production, confidence starts to erode.
That's the problem we built O2 to solve. And today, we're making it even easier to adopt -
O2 is now available as a GitHub Action on the GitHub Marketplace.
Web Applications in this digital era are relatively complex as it involves APIs, integrations, heavy workflows, etc.
Software companies are in a situation to buckle up complex releases frequently via CI/CD Pipelines. This rapid development has also resulted in the risk of bugs, and the margin of human error has increased.
Moreover, it has also resulted in performance as well as security issues and broken UI/UX experiences for the end user.
If you have been paying attention to Anthropic's tooling ecosystem over the last 12 months, Claude's web testing capabilities are genuinely exciting. The Computer Use API, the Claude in Chrome extension, and Claude Code's browser integration represent a real shift in what an AI model can do inside a browser - and the instinct to build a testing layer on top of it is completely rational.
Write your test cases in plain English. Let Claude drive the browser, find the bugs, and report back - no selectors, no assertions, no framework to maintain.
Add up every hour your engineers spent last sprint writing test selectors, debugging flaky CI runs, and patching broken scripts after a routine UI refactor. What does that number look like ?
Modern Software development is no longer just about writing code. It is about shipping software/product features faster than others without any bugs. With the Incorporation of agile methodologies, CI/CD pipelines, and microservices architectures, the SDLC cycle has drastically reduced, thereby enabling faster releases.
🧑💻 Vibe Coding + 🧪 Vibe Testing: Why Fast Code Still Needs Fresh Eyes
AI coding tools have dramatically changed how fast we can build software. UI refactors that used to take hours now take minutes. Boilerplate code is generated instantly.Documentation appears automatically.