# Weekly delivery review — App squad

Last 7 days, compared with the 7 days before. Source: DevStats MCP.

## Slack post

> **App — this week**
>
> **What moved**
> - 77 PRs merged (53 the week before), 63 opened. We merged more than we opened, so last week's pile (58 opened, 53 merged) is gone.
> - PR cycle time 1d 22h, down from 4d 20h. Deploy time did most of it: 54m, was 2d 4h. Review time 5h 20m, was 10h 46m.
> - 40 deploys, up from 20.
> - Cycle 53 is at 100% done (162 tasks). No aging issues.
>
> **What is stuck**
> - Pickup is now the slowest stage: 23h 41m of waiting before a first review.
> - #7539 (SUP2-965, DEMO - GitHub Copilot) is the oldest open PR at 1d 19h.
>
> **What we are doing**
> - First review the same day. Decide on #7539 today. Load the next cycle before the sync.

## Three actions for this week

1. **Cut pickup time.** Evidence: pickup is 23h 41m out of a 1d 22h cycle, more than coding (3h 47m), review (5h 20m) and deploy (54m 4s) combined. Review itself is fast once it starts, so the wait is about someone starting, not about rigor. Who: reviewers. Agree in the sync on a same-day first review and check the number next week.
2. **Merge or close #7539.** Evidence: SUP2-965 "DEMO - GitHub Copilot" is the oldest of the 3 open PRs at 1d 19h. #7545 (SUP2-962, 21h 57m) and #7547 (SUP2-969, 18h 28m) are inside a normal day. A demo PR that sits open becomes a long-lived branch. Who: the PR author and whoever is assigned to review it. Decide today.
3. **Load the next cycle before the sync.** Evidence: Sprint Progress shows Cycle 53 at 0% to do, 0% in progress, 100% done, while Daily WIP still shows 38 open tasks and 8 open branches (5 the week before). Work is in flight that the active cycle does not show. Who: me, before the meeting.

## One question the data cannot answer

Deploy time went from 2d 4h to 54m in one week. What changed? If it was deliberate (a pipeline fix, a release process change), we should write it down and keep it. If it was a one-off (a batch of PRs that had been waiting for one release), next week's number will bounce back and we should not celebrate yet.

## Notes on the data

- One week is a signal, not a trend. Issue cycle time (2d 18h vs 2d 20h) and PRs opened (63 vs 58) moved too little to read anything into.
- Aging Issues returned 0 items for the squad, so there is no cluster to name this week.
- All three open PRs carry SUP2 issue keys: the App squad's open review queue is currently support work.