build2 distinct publishers
Warp replays your stored agent runs to decide which model gets the next ticket
The replay only works on runs Warp's infrastructure already recorded, and correctness is graded by a judge model against a rubric you write. Warp's own 30-task bake-off cost $2,130.57. It finished in under four hours.
Reality
- Evidence41
- Adoption16
- Hype gap+33
- Incentives86