Agents

Code reviews with Buildprint

Buildprint Code Reviews analyze the changes between two Bubble branches and turn it into an organized review checklist. This helps you understand what changed, spot potential regressions, and build a test plan before merging.

Starting a code review

Navigate to Code Reviews and click Create. Select a source branch (the one with changes) and a target branch (usually test or staging).

Buildprint analyzes the changes and organizes the result into three sections:

  • Changes: Grouped summaries of what was added, removed, or modified across frontend, backend, workflows, and data model.

  • Potential issues: Likely regressions, risky behavior changes, assumptions, or product questions to confirm before shipping.

  • Tests: Actionable validation steps based on the actual diff - what to click, what data to check, and what edge cases to verify.

Each section shows items you can check off as you validate the branch.

Editing the review

The generated review is a starting point. Add, remove, or rewrite items as you work through the checklist. Common edits include:

  • Adding manual test cases the AI missed

  • Removing false positives from the potential issues list

  • Clarifying vague test steps with specific user flows

  • Grouping related changes together

Check off items as you complete them. This gives your team visibility into review progress.

Collaborating during review

Use the code review thread to chat with your team about edge cases, expected behavior, or concerns that come up during validation. This keeps technical discussion and QA coordination attached to the review itself.

Tag teammates to loop them into specific questions. The thread history becomes a record of why certain decisions were made or what was tested.

You can share a link to your review with anyone in your team.

Marking as merged

Once validation is complete and the branch has landed in the target, mark the code review as Merged. This closes the review and keeps your review history accurate.

Merged reviews stay in your history so you can reference them later - useful for tracking down when a regression was introduced or understanding why a change was approved.

Run code reviews on every branch that touches critical flows like payments, authentication, or user data. Small changes can have unexpected side effects in complex Bubble apps!

What code reviews catch

Code reviews are especially good at identifying:

  • Privacy rule changes that might leak data

  • Workflow modifications that skip validation steps

  • Database field deletions that break existing workflows

  • Conditional logic changes that alter user-facing behavior

  • API connector updates that change expected response handling

The AI can't catch everything - manual testing and human judgment are still essential. Use code reviews to organize and accelerate validation, not replace it.

Was this helpful?