Watch the recording
Open a recording from your library or the link returned by Rill. If it is still preparing, wait for Ready. Play the video and check that it shows the intended page and steps.
Pause at the important moment. Use a timestamped comment to explain what happened or ask a question. Record the flow again if you need to check a new version; reviewing an existing video does not run your app again.
Understand the browser details
Some recording methods include diagnostics: captured information from the browser. Use them to investigate what happened around a moment in the video.
| View | What to look for |
|---|---|
| Console | Messages and errors reported by the recorded page. |
| Network | Captured requests, status codes and failures. A failed request may explain a missing result. |
| Timeline | Captured browser events and interactions, when available. |
Chrome recording saves bounded console messages and request details, without request headers or bodies. Screen recordings and uploaded videos do not gain diagnostics after upload. Local-agent and cloud captures have their own capture settings. See what each method records.
An error may be unrelated to the task. Missing diagnostics do not establish that no error occurred. Use the video, the task and the captured details together.
Add the useful context
Recording: <Rill link>
Watch: <timestamp or short range>
Task: <what you tried>
Expected: <what should happen>
Observed: <what actually happened>
Version: <commit or build, if known>
Not checked: <what this recording does not cover>For a fix, record the same steps and test data before and after the change. Keep the two recordings clearly labelled. For an agent handoff, include the link and ask for observations grounded in the captured evidence.