Hands-On with KaneAI Beta: Web Testing Agent Features
KaneAI Beta has officially launched for testers and testing enthusiasts. It’s the world’s first end-to-end AI-powered software testing chatbot, packed with tools to help you test any web application quickly and efficiently. As a tester, I was eager to try out this innovative tool and finally got my hands on the beta version.
In this blog post, let’s explore the key features of KaneAI Web Agent, which looks very promising from a functional testing standpoint.
Intelligent Test Planner
When you open the KaneAI dashboard, you’ll see a text box asking for a test objective prompt. Describe the application you want to test and its functionalities. KaneAI will use this prompt to generate test steps and test cases and execute them on a cloud browser within the KaneAI Web Agent dashboard. The interface is very intuitive and can be easily used even by the non-technical folks.
Step-By-Step Prompting
Even though the intelligent test planner can create test steps, sometimes you’ll want to write test cases manually for accuracy. Using the step-by-step prompting feature, you can write a simple test case like “open google.com” and let the web agent process and execute it. Then you can add another step, like “search for sports shoes.” The agent processes and executes this in real time. You can pause, delete, edit test cases, and even add steps between existing ones.
Manual Interactions
The “Manual Interactions” feature captures your actions on the given cloud browser and converts them into test steps, which show on the left panel. This is helpful when you need to test actions that KaneAI can’t create on its own during test planning. It works similar to Selenium IDE, however its more versatile and user-friendly.
Conditional Statement
This standout feature lets you write conditional tests using If-Else statements in the step-by-step prompting approach, creating more control-driven tests. This is one of the most interesting feature as per my opinion, because during manual/automation testing, control-driven tests are frequently used so that the application is thoroughly tested.
Test Manager
KaneAI’s Test Manager tracks all executed and saved tests. It includes tabs for Insights, Test Cases, Test Plans, and Builds. You can launch, edit, and view the version history of your test cases here. This means that no need to use a separate application for tracking the test efforts and version control. Every testing process will be automatically captured in this intuitive dashboard.
Version Control System
Tracking different versions of test artifacts is crucial. KaneAI does this brilliantly, documenting changes to test scripts as new versions. This helps track modifications, increases productivity, and improves testing quality. You can compare versions and revert to older commits if needed.
Personal Opinion
KaneAI has the potential to disrupt the software testing industry. It offers a platform for testers to ease their tasks and build better-quality products. One of my favorite features is the clarity of the test steps written. The web agent provides detailed action types and descriptions, making it easier for stakeholders to understand what’s happening during execution.
Limitations/Drawbacks
Speed: The speed of creating and executing test steps on the cloud browser is slower compared to traditional automation tools. This might be a limitation of the Beta version. For few of the steps, I noticed that it took a lot of time. For incorrect prompts, the agent kept loading for a long time before throwing error, ideally it should have thrown error quickly to save the tester’s time.
Accuracy: The accuracy of the Web Agent in creating test cases using the intelligent test planner is minimal. The bot might not be advanced enough to understand the context of testing fully. The test cases created are very generic and straightforward and even quite similar for two completely different kinds of applications.
Lack of Browser Control Mid-Execution: After writing several test steps, I wanted to explore the website before adding more test cases. However, the tool doesn’t allow functional browser use like a normal tester would, which is a major drawback. Given the amount of time it takes to generate test cases, it becomes difficult to do frequent changes to the test suite.
Conclusion
Trying out KaneAI Web Agent Beta was definitely worth it. It’s amazing to see how AI and ML are transforming software testing. With such user-friendly tools, it’s safe to assume AI will play a significant role in the industry’s future. Testers must stay upskilled and ready to integrate these tools into their workflows.
