GitHub Action Planned
First-class GitHub Action that runs your scenarios against the PR preview deployment and posts the report as a comment.
Testing Suite · Coming soon
Same scenarios. Every deploy.
After every deploy, the scenarios you built for performance testing run against the deployed environment as smoke checks. If the homepage-to-checkout journey breaks, the pipeline halts and you can roll back. If LCP regresses past your budget on staging, promotion to production blocks. If a console error appears on checkout that wasn't there yesterday, you catch it before the next deploy reaches customers.
Testing Suite is on the roadmap. We'll write to you when there's a usable build.
CI run output mockup
1380 × 1035 · mockup
What it does
Build once. Reuse on every push.
Performance testing tells you whether your site stays fast under load. That's once a release, maybe weekly. The other 95% of the week, you ship code without testing what users actually see. Testing Suite is for that 95%.
After every deploy, it runs the scenarios you already built (homepage to checkout, search and add to cart, login and dashboard) against the deployed environment. Same real browser. Same isolation. Same Web Vitals capture. Smaller scale, faster runtime than a full performance test.
If a step in the scenario breaks, the pipeline halts and you can roll back. If LCP regresses past your budget on staging, promotion to production blocks. If a console error appears on the checkout page that wasn't there yesterday, you catch it before the next deploy reaches customers. You catch what unit tests miss.
And because the scenarios are the same ones you performance test with, there's one place to maintain them. No duplicated test code between your QA and performance pipelines.
What you get back
A PR check that links to the report.
A green check means the post-deploy scenarios passed and no Web Vital budget was busted. A red check means verification failed. Your pipeline can halt, roll back, or block the next promotion. Either way, the check links straight to the report for the run.
The report is the same one you get from a full performance test. Open any session. Watch the video. Read the logs. The bug that broke the build is one click away.
Roadmap
What's planned.
Here's the rough shape of what Testing Suite will ship with. We'll write to the waitlist as each piece lands.
GitLab CI Planned
A GitLab CI template that drops into your pipeline. Same scenarios. Same report. Run after every deploy to staging or production.
Threshold rules Planned
Set per-test budgets. LCP under 2.5s. INP under 200ms. No console errors. The check fails when budgets break, and your pipeline can halt or roll back.
Webhook API Planned
Trigger smoke tests from anything. Custom CI runners. Deployment hooks. Internal tools.
Diff reports Planned
Side-by-side comparison of the current run versus the previous green deploy. See what got faster and what got slower.
Slack notifications Planned
Failure alerts in a channel. Includes the broken scenario, the failed step, and a link to the session video.
What teams use it for
Six smoke-test patterns we hear on demo calls.
Different reasons to put the scenarios into your CI pipeline. Same forensic detail in every report.
Pull-request gates
Every PR with a preview deployment runs the critical-path scenarios against it. The PR check fails the same way a unit-test check fails, except this one catches Web Vitals regressions in a real browser.
Pre-release smoke
Tag a release candidate. The smoke suite runs against the staging environment. Ship only if it stays green.
Web Vitals budgets
Per-page LCP and INP budgets set at the URL level. The post-deploy check fails the moment a page busts its budget. Not later. Now.
Pre-merge diffs
Compare the candidate's preview-deployment run against the last green main-branch run. See exactly what got slower and what got faster, before the PR merges.
Regression hunting
When something regresses next week, look back at the smoke runs. The regression is in one of the deploys. The forensic detail tells you which.
Recurring nightly
Run the full smoke suite nightly against production. Catch slow regressions that ship-day smoke tests miss.