Ask: who is trying to do what, what could go wrong, and what would help them make a better decision? Pick a situation you can demonstrate without collecting sensitive data or testing someone else’s systems.

• Start now: choose the user and scenario. Define the safer outcome in one sentence.
• Build: build the smallest complete workflow.
• Test: test a normal case, a confusing case, and a failure case.
• By November 22: improve the roughest part, prepare your Devpost story, and submit before 5:00 PM Eastern.

• Does it explain why it flags something, rather than simply alarming the user?
• Can a user recover from a mistake or dismiss an incorrect warning?
• Does it collect less data than the problem requires?
• Is the main workflow understandable and accessible?
• What happens if the tool is wrong, offline, or unavailable?

Show the situation, the risk or uncertainty, the working solution, and the outcome. Then explain what you built during the event, what you reused, and one important limitation. Use fictional messages, sample identities, and test accounts.

No specific platform, paid service, or security background is required. Games, apps, websites, hardware, and low-code tools can all be effective. Use the Discussions tab to find teammates or ask general questions. For organizer support, use the contact link.