Purpose
- Reproduce bugs with a deterministic sequence of steps.
- Pinpoint the exact moment a regression occurs.
- Share replay evidence across teams.
- Build institutional knowledge of how elements behave over time.
Primary users
- QA engineers verifying fixes.
- Developers diagnosing UI regressions.
- Support teams attaching evidence to tickets.
- Product managers reviewing user interaction patterns.
Plan limits
Step replays quotas vary by subscription tier:| Plan | Step Replays/month | Retention |
|---|---|---|
| Free | 15 | 3 months |
| Starter ($99/mo) | 2,000 | 1 year |
| Pro ($199/mo) | 50,000 | 3 years |
| Enterprise | Unlimited | Custom |
Entry points
- Bug report detail view
- Project activity feed
- Element detail links to related replays
- Chrome Extension recording preview
Core replay flow
- Open a replay session from a report or element.
- Review the timeline and step list.
- Scrub the timeline to inspect state changes.
- Inspect console, network, and DOM context.
- Share or export replay metadata.
CI workflow (GitHub Actions)
Step Replays now include a CI workflow so QA can move from “recorded issue” to “PR-visible validation” without manual handoff.CI workspace
- Open each replay in a dedicated CI workspace.
- Choose a validation profile (
smoke,regression, orauth). - Trigger
Run in CIdirectly from the replay context. - Track run status and history from the same screen.
Project-level CI configuration
- Configure GitHub repository mapping per project.
- Select default workflow file and branch/ref target.
- Set allowed branches to limit where replay validation can run.
- Optionally require approval for protected branches.
Replay artifact bundle
Each CI run can consume a generated replay artifact bundle that includes:- Playwright replay test scaffold
- Replay and step metadata
- Comments and step log data
- Screenshot manifest and available screenshot files
- Optional rrweb event log
- PR comment draft markdown
Run lifecycle tracking
Replay validation runs are tracked from dispatch to completion with statuses such as:QueuedRunningPassed/FailedDispatch failed/Policy blocked
Page structure (web app)
1. Replay header
- Session metadata (URL, user, timestamp)
- Status and environment badges
- Share and export controls
2. Playback viewport
- Recorded UI playback with controls (rrweb-based)
- Speed selector (0.5x, 1x, 1.5x, 2x) and pause/resume
- Fullscreen mode for detailed inspection
- Screenshot fallbacks for missing frames
3. Timeline and steps
- Interactive timeline with event chips (click, scroll, focus, mutation, navigation, error)
- Playhead scrubbing with drag support
- Step list panel with timestamped events
- Auto-scroll to active event during playback
4. Diagnostics panel
- Console logs with severity filters
- Network requests summary
- DOM and selector context snapshot
- Error highlighting with stack traces
Timeline event types
The timeline captures and visualizes these interaction types:- Click: Mouse clicks and double-clicks
- Focus/Blur: Element focus transitions
- Scroll: Page and element scrolling
- Mutation: DOM changes and UI updates
- Navigation: Page navigations and URL changes
- Error: Console errors and exceptions
Replay triage taxonomy
Replay comments and timeline pills use the CRO issue taxonomy:- Broken Experience
- UX Friction
- Unclear Messaging
- Weak CTA
- Trust Gap
- Form Friction
- Performance/Speed
- Mobile Usability
- Tracking Gap
- No Issue
Data captured
- Event timeline and interaction metadata
- DOM snapshots and element references
- Console logs and error stack traces
- Network request summaries
- Element selector stability data
- Optional screenshots or frames
Key actions
- Jump to a step or timestamp
- Copy element selectors from a replay
- Export logs and share links
- Add comments/annotations at specific timestamps
- Tag events for team review
- Run replay validation in CI
- Download CI artifact bundle
Integration with Element Library
Step replays are connected to the Element Library:- Element captures link to related replays showing the element in action
- Selector stability is validated against replay data
- Bug reports can reference specific replay timestamps
Known limits
- Cross-origin frames may reduce detail
- Large sessions may stream data progressively
- Retention periods vary by plan (3 months to custom)
- Maximum recording length: 30 minutes per session
Export capabilities
Step replays can be exported as reproduction steps for issue tracking and test automation:- Jira: Formatted reproduction steps with console logs and environment details
- Linear: Issue-ready markdown with technical context
- Playwright: Auto-generated TypeScript/JavaScript test code