Submit
Register your build for the showcase. Update the demo URL any time — resubmitting updates your entry, never duplicates it.
Or ask your agent (once armed):
Submit our project. It's called <name>, deployed at <your worker URL>, and here's what it
does: <one sentence>.
If your agent can't reach the MCP server
Use the form above instead, or ask a host — nothing about the showcase depends on the MCP server being available.
How judging works
Final score = agent rubric (reviewed and approved by a human host) + room vote.
| Dimension | What it asks |
|---|---|
| Works | Does the deployed agent actually respond and do the thing? |
| Agentic depth | Tools, state, scheduling, durability — or a completion call in a trench coat? |
| Industry relevance | Would someone in this industry recognise the problem as real? |
| Path to production | Is there a credible route from this demo to something shippable? |
| Human approval gate | Does your agent have a human approval gate? |
A Workflow scores your submission automatically — screenshot, a probe if you registered
handle_request, a rubric pass — then pauses for a human host to approve or override it
live, on the big screen, before it counts. Same accountability theme, aimed at the
judging itself.
At the showcase
Each team gets a couple of minutes live on the Mission Control board. Then the finale fans
one cross-industry scenario out across every team that opted into handle_request — if you
built it, bring it.
Good luck.