Why the format matters, not just running a retro at all
A retrospective run with the wrong format for the moment tends to produce the same generic feedback every time - a team that just had a rough sprint needs a format built for surfacing frustration (Mad/Sad/Glad), not a neutral Start/Stop/Continue that treats a bad sprint like any other.
How the picker works
Three quick questions - how the last sprint went, how much time you have, and whether the team is remote or co-located - map to a format proven to fit that situation. No AI, just a decision tree matching common retrospective-facilitation guidance.
Using the result
Every format ends with the same real requirement: concrete actions with owners, not just a list of feelings. A format is a way to surface what needs discussing - the retro only pays off if something changes afterward.