Been doing QA at an agency for a while, and one thing that eats up more time than it should is cleaning up half-written bug notes before they’re actually ticket-ready: vague descriptions, missing repro steps, no build info. Got tired of it, so I built a small AI-assisted tool that takes raw notes like:
“found a bug on edit profile ios, long username overlaps the dob field, build 2.3.1, iphone 15, ios 17.4”
and turns that into a fully structured ticket: title, description, a note pointing devs at the likely cause, steps to reproduce, expected/actual result, environment, all filled in properly. If something’s missing, like the build number, it asks instead of guessing rather than making something up.
Sharing it free for anyone who wants to try it. Happy to share a risk-based test prioritization tool I built off the ISTQB framework too, if that’s useful to anyone.
Genuinely curious though: is this a pain everyone deals with, or mostly a me problem? How do the rest of you handle turning rough tester notes into something ticket-ready — do you have your own shortcuts, templates, or tools for it, or is it just manual cleanup every time?
the cleanup is real, but i’d separate formatting from evidence. ai can turn rough notes into a cleaner ticket, yet it cannot recover the starting state, ordered steps, or first wrong transition if the tester never captured them.
i’d make the workflow: start capture before reproducing, preserve expected versus actual plus the browser path, then let ai structure the ticket and ask about anything still missing. mark generated “likely cause” text as a hypothesis so engineering can distinguish observation from inference.