Suggest
A cleaner title, in one press
Reads the report body and produces a short, searchable title that stands up in a queue of two hundred.
Pay button not clickable on iOS viewports under 400px
AI triage
Turn on AI assistance and the inbox will suggest cleaner issue titles, rewrite rambling reporter notes into clear reproduction steps, and translate reports from any language back into yours. Off by default; scoped per workspace; feedback text sent for processing only, never for training.
Three helpers
AI never runs unbidden. It sits behind buttons on the issue and only acts when you press one.
Suggest
Reads the report body and produces a short, searchable title that stands up in a queue of two hundred.
Pay button not clickable on iOS viewports under 400px
Rewrite
Turns a messy reporter note into a clear description with reproduction steps, expected and actual behaviour.
On Safari iOS at 390 × 844, the pay button overflows its container and cannot receive tap events…
Translate
Translate title, description and comments into English or the workspace language. Original text kept alongside.
“Der Pay-Button lässt sich auf dem iPhone nicht anklicken.” → “The pay button can’t be tapped on iPhone.”
Before & after
A short reporter note becomes a triage-ready issue in a couple of clicks. Both versions are kept — the original stays on the issue so nothing the reporter said is lost.
Original
“hi the pay thing looks weird on my phone can’t click it, sarah”
Suggested title
Pay button not clickable on iOS viewports under 400px
Rewritten description
On Safari iOS at 390 × 844, the pay button overflows its container and cannot receive tap events. Reproduces on the checkout step of the pricing page.
Controls
Every workspace starts with AI disabled. An owner has to enable it, and it can be turned off again at any time from the same settings page.
One switch enables all three helpers; disabling removes the buttons from the UI immediately.
Restrict who can trigger AI actions — everyone, admins only, or nobody until you’re ready.
Feedback text is sent to the model provider for inference only. It is never retained or used for training.
Every AI action appears in the audit log with actor, action and the accepted output — same as any manual edit.
Enable AI triage per workspace, keep the audit log clean, and revert anything you don’t like.