There are two doors, and picking the right one saves you a day. Anything the Mac app did or failed to do should be reported from inside the app — it can send its own logs, and we can't guess at them from here.
A run that stopped, a photo or voice memo it didn't read, a chapter that came out wrong, the app not launching — all of it lives in logs on your Mac. The app can bundle those and send them with your description in one go.
You can see exactly what gets sent before it goes.
In the same sheet, tick Work out what went wrong, on this Mac and a Claude Code session on this Mac looks into it and sends us a second report, linked to yours, which you do not see first. It is often faster than waiting for us.
A question before you install, an idea, or you can't get the app to run at all. This reaches one person, so plain English is fine.
Usually one of three things: your Mac slept, a login expired, or the AI hit a usage limit and parked itself. The app's own status line says which. Nothing is lost either way; your archive survives being interrupted, and the next run picks up.
Some sites can't be automated at all, and that's expected rather than broken. Tell us which one, and what you were trying to do.
Almost always the install step. This is the part we most want to hear about, so write here rather than fighting it. What it needs, and why.
Correct it in the app rather than telling us; corrections are remembered for every future chapter. Tell us only if it keeps making the same mistake after you've corrected it.