qc vrt
Run visual regression tests against Quant-hosted projects. Uses Playwright and Chromium to capture screenshots and Pixelmatch to detect differences between the Quant-served pages and their remote origins.
For an end-to-end walkthrough including suggested thresholds and CI integration, see the visual regression testing guide.
Synopsis
Section titled “Synopsis”qc vrt [options]Description
Section titled “Description”By default the command tests every project in the active organization. Pass --project to limit it to one or more named projects. Output is a set of screenshots and (optionally) a CSV report.
Options
Section titled “Options”| Flag | Description | Default |
|---|---|---|
--project <project> |
Specific project(s) to test. Comma-separated for multiple. | All projects in the organization. |
--threshold <threshold> |
Pixel difference threshold, between 0 (identical) and 1 (any difference). |
0.01 |
--max-pages <maxPages> |
Maximum pages to crawl per project. | 10 |
--max-depth <maxDepth> |
Maximum crawl depth. | 3 |
--csv <file> |
Path to write the CSV report. | — |
--quant-auth <credentials> |
Basic auth credentials for Quant URLs (user:pass). |
— |
--remote-auth <credentials> |
Basic auth credentials for the remote/origin URLs (user:pass). |
— |
--output-dir <dir> |
Output directory for screenshots. | ./vrt-results |
Examples
Section titled “Examples”Test a single project
Section titled “Test a single project”qc vrt --project my-projectTest multiple projects with a strict threshold
Section titled “Test multiple projects with a strict threshold”qc vrt --project site-a,site-b --threshold 0.01Limit crawl scope and export CSV
Section titled “Limit crawl scope and export CSV”qc vrt --max-pages 50 --max-depth 2 --csv vrt-report.csvTest a password-protected staging site
Section titled “Test a password-protected staging site”qc vrt --project staging --quant-auth admin:secret --remote-auth admin:secretSee also
Section titled “See also”- Visual regression testing guide — detailed setup and CI patterns.
qc crawler— crawl content into a project.
