Artifact Vault
Browse, debug, and manage screenshots, videos, and Playwright traces from failed test runs.
Overview
When a test fails, your CI runner captures evidence — a screenshot of the page at the moment of failure, a video recording of the whole run, and a Playwright trace file containing every network request, DOM snapshot, and action. Without a central place to browse this evidence, engineers spend time hunting through CI artifact downloads or S3 buckets.
Artifact Vault is that central place. It organises all failure evidence by test run, lets you view screenshots in a lightbox, step through Playwright trace timelines, and manage how long each type of artifact is kept before it expires.
Three artifact types
Artifact Vault tracks three types of test evidence: Screenshots (PNG captures at failure), Videos (MP4 recordings of the full test run), and Traces (Playwright .zip files with network, DOM, and action replay data). Each type has its own default retention policy.
Quick Start
- Navigate to QA → Artifact Vault in the sidebar.
- The screen opens on the Visual Debugger tab with all test runs listed. Filter by status using the All / Failed / Flaky tabs.
- Click a test run row to expand it and see individual test artifacts.
- Click a screenshot thumbnail to open the full-size lightbox.
- Switch to the Trace Viewer tab to step through the Playwright action timeline for a specific run.
- Switch to the Retention tab to review expiry dates and adjust per-file policies.
The Three Tabs
Visual Debugger
The default tab. Lists all test runs as collapsible accordion rows. Each row shows:
- Test file and suite name
- Run status badge (Failed / Flaky / Passed)
- Artifact count (screenshots, videos, traces)
- Timestamp of the run
Expanding a row reveals a gallery of screenshot thumbnails. Click any thumbnail to open the full-size lightbox. Use arrow keys or the on-screen chevrons to navigate between screenshots in the same run.
A Video Capture section below the screenshot gallery provides a playback link to the MP4 recording. Videos open in the browser's native video player — no plugin required.
Trace Viewer
Displays a horizontal timeline of every action Playwright took during a test run. Each step in the timeline shows:
- Action type (click, fill, navigate, waitForSelector, etc.)
- Target selector or URL
- Duration in milliseconds (shown as a proportional bar)
- Status: success (green), warning (amber for slow steps >500ms), error (red for failures)
The timeline is useful for pinpointing exactly which step timed out or threw. A step with a 3-second duration bar that precedes an error is almost always the root cause of the failure.
Compare across runs
Run the trace viewer on a passing run and a failing run side-by-side (open two browser tabs) to spot where the timing diverges. The duration bars are the clearest signal.
Retention
Shows a storage overview panel and a per-file retention list. The storage panel breaks down total disk usage by artifact type. The per-file list shows every artifact with its expiry date and a countdown indicator:
- >7 days — Plenty of time. No action needed.
- 3–7 days — Expiring soon. Download if you still need it.
- <3 days — Expiring imminently. Download or extend now.
Click the Extend button on any artifact to reset its retention timer. Click Delete to remove it immediately and free up storage.
Default Retention Policies
Each artifact type has a different default retention window, reflecting its storage cost and typical usefulness:
| Artifact type | Default retention | Typical size | Reason |
|---|---|---|---|
| Screenshots | 30 days | 50–200 KB each | Small and highly useful for async bug reports. Kept longest. |
| Videos | 14 days | 2–20 MB each | Large files. Useful for two sprint cycles to cover retros and reviews. |
| Traces | 7 days | 0.5–5 MB each | Most useful immediately after a failure. Engineers investigate within days. |
These defaults can be overridden per file in the Retention tab, or globally in Settings → QA.
Status Filters
The filter bar at the top of the Visual Debugger tab has three options:
- All — Show every test run regardless of status, including passed runs that have artifacts from a previous failure.
- Failed — Show only runs where the test failed on its final attempt. These have the most actionable artifacts.
- Flaky — Show runs where the test passed only after retries. Screenshots and traces from the failed attempts are preserved even though the run ultimately passed.
Storage Management
The Retention tab's storage overview shows total usage and a doughnut-style breakdown:
- Screenshots — Usually the largest category by file count, smallest by total bytes.
- Videos — Largest total bytes. Set a shorter retention window or disable video capture on low-priority test suites to control storage growth.
- Traces — Moderate size. The most diagnostically dense artifact type per byte.
- Expiring Soon — Count of artifacts expiring within 7 days. Acts as a reminder to download anything you need before it is auto-deleted.
Storage grows fast with video
A CI pipeline running 500 tests per day can accumulate several gigabytes of video data per week. If storage is a concern, reduce the video retention window to 3–5 days or disable video capture for suites that have good screenshot coverage.
FAQ
.zip format that can be opened with npx playwright show-trace..zip trace file and run npx playwright show-trace trace.zip in your terminal. This opens the Playwright Trace Viewer in your browser with full interactivity — network requests, DOM snapshots per action, and console logs.