Release notes··4 updates

Self-improving PR reviews, GitLab CI attribution, and plan defaults in chat

4 updates shipped on September 1, 2026.

  1. Highlight

    Featured this release

    GitLab CI run attribution

    Runs triggered from GitLab CI show GitLab icon with linked branch and commit, self-hosted included.

    Runs triggered through the public API from GitLab CI are now recognised as first-class GitLab runs, with branch and commit links back to GitLab, support for self-hosted GitLab instances, and GitLab as a filter option alongside your other trigger sources.

    More on this: the GitLab integration and PR and CI/CD testing.

    # Direct link

  2. Highlight

    Featured this release

    PR reviews learn from your feedback

    A thumbs-down with a written reason now shapes the next review on that project.

    When you rate a PR review down and explain why, QA.tech turns that into a lasting improvement for future reviews on the same project – either a standing rule the next review is briefed on, or a memory it recalls when relevant.

    Outdated rules get narrowed or retired as your app changes, so the reviewer stops flagging things that no longer apply.

    More on this: pull request testing and why developers should not test their own code.

    # Direct link

  3. Update 03 of 04

    Full run results on the PR review page

    PR reviews now show the same List and Dependencies views as the results page.

    Each PR review card renders the full run results view for that round – switch between List and Dependencies to see how the tests relate, then follow "View Full Run" to open the complete run on the results page.

    More on this: pull request testing and PR and CI/CD testing.

    # Direct link

  4. Update 04 of 04

    Pin test plan defaults from chat

    Set a test plan environment, device and agent in chat and it sticks for every future run.

    Chat could already create and edit test plans; the default environment, device preset and agent lived only in plan Settings. Now you can pin all three directly in chat when creating or editing a plan, and ask for a plan's current defaults without leaving the conversation. Scheduled and manual runs use whatever you set.

    More on this: multi-environment testing and smoke testing.

    # Direct link

Your team moves fast. Can your testing keep up?

QA.tech agents test your product autonomously, so moving fast never means shipping broken. See how it works in a 30-minute demo.

Get a demo