Inspiration
As the year comes to an end, many users look forward to year-long wrap ups, like Spotify Wrapped or Apple Music Rewind. For us mere programmers, the only thing we have is that snarky little green contribution graph year round. We wanted to change that and provide a truly unique and fun experience to visualize Git repos in a different light.
What it does
Our app combines data from GitHub/Git, static code analysis on repositories, and open AI models to create CodeWraps. These are bite-sized code insights that go beyond what IDEs or GitHub typically provides, primarily for entertainment, as to what type of programmer(s) are part of a repo. Are you a comment queen or afraid of them like the plague? How deep (and horrifying) is your most deeply nested loop? What would a poem using words from your source code sound like? There's only one way to find out.
How we built it
The front-end is built in Flutter. We chose this due to the need/want to imitate the beautiful animations that Spotify Wrapped has, which was our main inspiration for this project. As well, it is easy to deploy a Flutter project to the web if need be. The back-end is written in Flask, but calls GitHub and other APIs.
Challenges we ran into
- The front-end takes a long time to code per screen.
- Authorization for API calls is not fun. This is also the reason we do not have a live working copy, as we did not want to leave an exposed API with sub-optimal security at this moment.
- We were forced to reduce the scope of the application due to time constraints
- Not enough artistic talent! We had originally planned to "paint a picture" of each repo, inspired by the insights. However, we did not have the artistic ability nor the time to create all the assets in a way that would work together.
- Buggy bugs, as always.
Accomplishments that we're proud of
- Although the backend and frontend don't connect at this time due to API authentication issues, we worked hard on them separately and they do perform once connected. We are proud of the progress we made on both of them.
- The pages we designed came together nicely in the frontend, and we had an articulate, clean, and sleek model to work off of which greatly benefited us in the short period of time.
What we learned
- We all got to work with technologies that we were unfamiliar with and stepping out of our comfort zones in this short period of time was an exciting challenge.
- As well, we learned the importance of allotting a time limit to different portions of the project lifeline, because we found that generating and designing an idea takes a lot more time than anticipated.
What's next for CodeWrap
- Resolve API authentication issues, so as to connect the front and backend.
- There is still development left to make it a fully functioning prototype
- We would like to implement the feature to generate a painting based on a combination of user characteristics such as their most commonly used variable name. Ai technologies that can accomplish this task already exist, so we would like to integrate them with our app. This will liven the user's experience.


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