Liscense Guard

Inspiration

Many people will go through liscense process, and not many would exactly know what liscense they need. I built a tool to help with that.

What it does

How we built it

Built with Next.js and Typescript and deployed with Railway.

We used public and usuable(in terms of liscenses) repos to test it out(using GitHub API). It reads the repos' package.json to check dependecies list and liscenses. We seperated most used lisceses by how contagious it is to others. For example, GPL-2.0 and Apache-2.0 didn't work together. For this part, we didn't use a LLM to determine because a simple hard coding worked better. AI could make mistake.

We used AI to explain what happened with liscenses about the repo in plain text so that the user could understand.

Challenges we ran into

It was to decide what to make an AI do and what not to. At first it felt like letting an AI do liscense determination would work better and faster, but we realized that it would work becasue it could lead to two different answers from a repo.

Accomplishments that we're proud of

We think it was about how efficient and effective the AI usage here was. We thought about dependency on LLM capability and knowledge.

What we learned

We learned that open source liscensing isn't about a preference. We were curious about the open sourcing process before we actually came up with idea and thought that this could be what our project could be about.

What's next for Liscense Guard

The first thing we should do is work with "transitive dependency". We just don't have that yet.

Built With

• github-api
• next.js
• node.js
• railway
• react
• tailwindcss
• typescript