ContribAI

ContribAI helps beginners start in open source by finding the right GitHub issue, breaking it into simple steps, planning work around free time, and guiding them to a PR.

Comment

Inspiration

Open source is exciting, but getting started is hard. Many contributors do not know which issue to pick, how difficult it will be, which files to look at, or how to plan their work around their free time. We built ContribAI to make open source feel less confusing and more guided. Our goal was to create a smart teammate that helps users choose the right issue, break it into simple steps, and stay on track until they are ready to submit a pull request.

What it does

ContribAI helps people contribute to open source in a more organized way.

It lets a user sign in, enter a GitHub repository, and analyze the repo’s open issues. It groups issues by tech stack, sorts them by difficulty, and recommends the best issues based on the user’s availability and selected path.

After the user picks an issue, ContribAI creates a simple action plan. It explains the issue in easy words, suggests relevant files to check, breaks the work into sessions, and tracks progress as the user moves forward.

It also helps with real contribution workflow tasks like:

checking GitHub progress suggesting next actions helping when the user gets stuck drafting a maintainer comment drafting a pull request turning the plan into Google Calendar events giving voice support for accessibility

How we built it

We built ContribAI as a full-stack web app using Next.js 15, React 19, and TypeScript.

For authentication, we used Auth0 so users can securely log in.

For AI features, we used Google Gemini 2.5 Flash through the @google/genai SDK. Gemini helps us create issue plans, give coding help, generate progress updates, rewrite schedules, and draft maintainer comments and PR text.

For repository data and live contribution tracking, we used the GitHub API. This lets us fetch repository issues, inspect the repo tree, detect branches and commits, and check whether a pull request has been opened or merged.

For scheduling, we used the Google Calendar API through the googleapis package, so planned work sessions can be added directly to a calendar.

For accessibility and voice output, we used ElevenLabs text-to-speech.

Overall tech stack:

Next.js React TypeScript Auth0 Google Gemini API GitHub API Google Calendar API ElevenLabs API

Challenges we ran into

One challenge was turning raw GitHub issue data into something actually useful for a beginner. Issues are often vague, inconsistent, or missing labels, so we had to build logic to estimate stack, difficulty, and time in a practical way.

Another challenge was making the AI output structured and reliable. Since we needed plans, comments, and PR drafts in usable formats, we had to carefully design prompts and add fallback logic when AI responses were imperfect.

We also had to connect multiple services together: authentication, GitHub data, Gemini responses, calendar scheduling, and voice output. Making all of these work smoothly in one user flow took a lot of debugging and iteration.

Accomplishments that we're proud of

We are proud that ContribAI is not just a chatbot. It supports the full contribution journey.

It can:

help users choose a repo issue create a realistic work plan split the work into sessions reschedule if the user gets interrupted track contribution progress on GitHub help communicate with maintainers prepare a PR draft add sessions to Google Calendar We are also proud that we included accessibility support with voice output, which makes the experience more inclusive.

What we learned

We learned that good developer tools are not only about AI, but about workflow. AI becomes much more useful when it is connected to real systems like GitHub and Google Calendar.

We also learned the importance of strong prompt design, fallback handling, and clear user flow. A smart feature is only helpful if the result is simple, reliable, and easy to act on.

Most importantly, we learned that beginners need guidance in small steps. Breaking a contribution into manageable sessions can make open source feel much less intimidating.

What's next for ContribAI

Next, we want to make ContribAI even more personalized and powerful.

Our next steps include:

smarter issue matching based on user skills and history deeper GitHub integration for branch and PR creation better team collaboration features for hackathons and student groups support for more calendars and productivity tools stronger accessibility features improved AI feedback on code changes and testing Our long-term vision is for ContribAI to become a full contribution companion that helps more people confidently join and stay in open source.

Built With

Share this project:

Updates