Introduction to tests
Buildprint Tests use an AI agent and a real browser to test your Bubble app. You describe the user journey in plain English, choose a Bubble branch, and can review what the agent did at each step.
How tests work
Each test is a visual flow that starts with its browser and user settings, then follows connected steps.
You choose the Bubble branch and AI model.
Buildprint opens the app in a browser.
The agent signs in as the assigned test user when one is configured.
The agent follows each instruction and condition in the flow.
Buildprint records the result, timing, screenshots and other evidence.
The agent works through the app like a user. It can navigate pages, enter values, click controls, upload files and check visible results.
Parts of a test
Tests are built from a small set of reusable parts.
Test: a complete user journey that you can run.
Test step: an action or check written in plain English.
Condition: a decision that sends the agent down a different path.
Component: reusable steps, such as a login flow.
Test user: a Bubble user account the agent can use during a run.
Run: one attempt to complete a test against a selected branch.
What to test
Tests work best for important user journeys with a clear expected result.
Signing up, logging in and logging out.
Creating, editing or deleting records through your app.
Submitting forms and completing multi-step flows.
Checking permissions for different types of users.
Confirming that important pages and controls still work after a deployment.
Branches and data
You can run tests against Live, Test or a custom Bubble branch.
Live runs use your live app and database. Test and custom-branch runs use development data. Assign separate users for Live and development runs when the test needs to be logged in.
A test can perform real actions in the selected database. Use dedicated test accounts and check the selected branch before running against Live.
Test results
Each run receives an overall status based on its steps.
Passed means the test completed successfully.
Warning means the run continued after finding a problem.
Error means the test failed.
The Dashboard shows recent runs and regressions. History keeps earlier runs so you can inspect evidence and compare results over time.