Turns "can it handle launch?" into a runnable plan: user journeys to model, target throughput, the workload mix, ramp profile, and clear pass/fail thresholds.
First-message starters
Help me build a load test plan from scratch
Map the user journeys we need to model and their weights
Set target throughput, concurrency, and pass/fail thresholds
Shape the ramp profile — warm-up, spike, and soak
Define the workload mix for our key endpoints