Bug reports with annotated screenshots
The bug report that says "the invoice looks wrong" is missing the one thing support needs: the screen. Asking users to screenshot, crop, annotate and attach it themselves loses most of them at step one — and the reports that do arrive carry no ticket id, no user, no context your tooling can route on.
The capture matches what they saw
Computed-style capture means their dark mode, their data and their half-scrolled table come through as rendered — which for a bug report is the entire point.
The markup is where the information is
Arrows, boxes, highlighter, text and a caption box — the caption is usually the sentence that explains what they were actually complaining about. Redaction genuinely destroys the pixels underneath, so a screenshot of a billing page can be shared safely.
It arrives as data, not an attachment
Attach params at the call site — user id, ticket id, build number — and the webhook delivers file URLs plus that context, signed. Routing a report to the right queue is a lookup, not an email parse.
Questions people ask
- Does anything upload before the user confirms?
- No. Capture, annotation and redaction all happen in the browser; nothing leaves the page until the user hits Send, and redacted pixels are destroyed before the file is even encoded.
- How does the report reach our tracker?
- Your endpoint receives a signed webhook with the file URLs, the caption and your params. From there it is one handler to create the ticket in whatever you use.