๐ Astro Aura ๐
โจ Inspiration โจ
We wanted to solve the mental and physical health problems of the effect of isolation on astronauts. We wanted to create a fun and engaging way for astronauts to stay active and sane while also socializing with their fellow astronauts while on a long mission.
โ๏ธ What it does โ๏ธ
When an astronaut signs up, their name and mission duration is stored. We then used a Large Language Model (LLM) to generate an activity plan for them to follow.
Every day, an astronaut is given 2 missions to do.
- One mission is part of their overall plan
- For every day of a month, the astronaut is encouraged to do a different activity every day as outlined in their overall plan.
- If the astronaut fails to complete this task, they receive a deduction from their aura.
- Daily challenge
- All astronauts have the same daily challenge, and they're awarded with more aura than the above task upon completion
- Since these tasks are optional, the astronaut won't lose aura if they fail to complete them.
Completing challenges earns aura, which is our take on a points system. The astronaut with the most aura is the coolest person on the spaceship. Be careful though! Missing a day of your planned tasks will make you lose aura.
๐๏ธ How we built it ๐๏ธ
Our app stack includes Supabase for the database and Next.JS for full-stack.
- We provisioned the table in Supabase.
- We designed the system prompts, purchased OpenAI API credits, and wrote the
gpt-4o-miniintegration. - The backend team wrote the API routes while the frontend team planned out the frontend in Figma.
- After the design was done, we added user authentication.
- We focused on implementing the frontend, provisioning work among the members.
- Then, we linted the app and improved the overall code style.
- ๐๐๐ using Vercel.
๐งจ Challenges we ran into ๐งจ
Unfortunately, Astro Aura was not the first idea we came up with. We scrapped our original idea a few hours into the hackathon, ultimately losing 2.5 hours of hacking time.
We also ran into several merge conflicts along the way, as well with issues with Server-Side Rendering (SSR) and Node.js' fetch library link system.
๐ Accomplishments that we're proud of ๐
- โ We shipped the app successfully
- โ We learned a lot about system prompt design and ultimately came up with a good system prompt
- โ We successfully added user authentication to our app faster than we expected
๐ What we learned ๐
- OpenAI's API and prompt engineering
- Refreshed knowledge of Supabase and SQL
๐ What's next for Astro Aura ๐
"Credit" and "streak saver" system as micro-transactions- Syncing data across missions
- Multiple daily challenges
Built With
- nextjs
- openai
- react
- supabase


Log in or sign up for Devpost to join the conversation.