Submissions
How the field was judged.
Every submission was scored against the same rubric: memory load-bearing 40, innovation 25, execution 20, pitch 15, then a PMF bonus and a verified partner-stack multiplier.
judging closed · 92 teams submitted
What a finished submission included
Four things. Every scored submission carried all of them.
Public repo
Clean commits and a README with complete setup and run instructions.
2 to 5 minute demo
Show the problem, the product, how it works, and a fresh session recalling earlier state.
Team & partner stacks
Name the builders and every Base or Virtuals stack used.
Memory implementation note
Explain what your agent persists, recalls, and uses to make decisions.
One rubric · one leaderboard
How judging worked
Projects cleared a pass/fail gate, then scored against four weighted criteria. A genuine PMF bonus adds up to 10, and verified Base and Virtuals stacks then apply the multiplier.
100-pt rubric + 10 PMF bonus = 110 max, before multiplier
- Memory is load-bearing 40
- Innovation & originality 25
- Technical execution 20
- Pitch & presentation 15
- PMF bonus +10
Memory is load-bearing · 40%
How central and sophisticated the memory use is. Recall is competitive; coordination and dynamic-storage patterns top the band.
Innovation & originality · 25%
A novel idea and real usefulness. Clever but useless caps out low.
Technical execution · 20%
Clean and robust, no smoke and mirrors. Survives a second run and a curious judge.
Pitch & presentation · 15%
A tight 2 to 5 minute story where the load-bearing moment is unmistakable.
Verified partner stacks then multiply the weighted score, up to x1.25. See how the Builder Score worked →
The board is final.
92 teams submitted, 38 were code-reviewed, and twenty took an award.
See the results →