No more babysitting the brief queue when a generation run dies mid-flight.
New
- Automatic recovery for stuck briefs — Briefs that get stuck in drafting because the LinkedIn or blog generation pipeline crashed mid-run are now auto-recovered. An hourly background sweep flips any brief that has not moved in 15 or more minutes back to the queue and posts a notification telling you exactly which briefs were recovered, so you no longer have to click Release by hand.
Improved
- An empty AI draft is now treated as a real failure — The Generate with AI flow in the LinkedIn and Blog composers no longer silently saves a blank article. The brief is released back to the queue and you get a clear "AI generation returned an empty draft" message instead.