All data captured
Multiplayer feeds full-stack, unsampled, pre-correlated data to your coding agent, by default. After following the quickstart steps you can expect the following data to automatically be correlated per session.
Frontend data
- User clicks
- Page navigations + loads
- Session metadata
- Browser information
- OS information
- Device type
- Screen size
- Pixel ratio
- Cookies enabled/disabled state
- Harware concurrency
- Package version
- DOM events
- Console messages (message + stack trace)
- Network requests
- HTML source code
ℹ️ For a detailed overview guide, please refer to: How Multiplayer records and replays user interactions and data.
Please note that Multiplayer is built with privacy in mind and, by default, all user inputs are masked and we do not capture <canvas> content by default (although it can be enabled, see here).
Backend data
- Backend errors
- Correlated distributed traces and logs per session, with no sampling
- Request / response content per session, including from inside the system components (e.g. going from one service to another)
- Header content per session, including from inside the system components (e.g. going from one service to another)
ℹ️ For a detailed overview guide, please refer to: How Multiplayer backend tracing works
🔑 Pro tip: Multiplayer is backend agnostic and supports OpenTelemetry: you can seamlessly integrated it with whichever observability stack you already use.
Additional context
Before the data reaches the coding agent, we automatically also add release context:
- Build information
- Deployment timestamps
- Recent commits
- Version of each service involved in the failure
Developer can also add the following for each recording through the web dashboard:
- User comments / feedback
- Team annotation (sketches on top of the recordings or notes per data point)
Next steps
🚀 If you’re ready to trial Multiplayer with your own app, you can follow the steps in the quickstart.
📌 If you have any questions or want to report bugs please open a GitHub issue! 💜