How to spot rising open source projects

The tools everyone talks about next year are usually sitting on GitHub today with a few hundred stars and a rough README. Finding them early lets you shape them with feedback, contribute while the codebase is still small, and adopt them before your competitors do.

The hard part is not finding new repositories. Thousands appear every day. The hard part is filtering them down to the handful that matter to you.

Look at speed, not size

Total stars mostly measure age and marketing. A two year old project with 8,000 stars and a two week old project with 2,000 stars are not in the same situation.

The more useful number is stars per day since creation. It tells you how fast attention is arriving right now. A young project gaining stars quickly is being shared, bookmarked, and discussed, which is exactly the signal you want when you are looking early.

That is how our Rising on GitHub page ranks by default. It looks at projects created in the past month and sorts them by how fast they are being starred.

Narrow to your stack first

Momentum is only interesting if the project could end up in your work. Filter by the language you actually ship in before you read a single README. A fast rising Rust database is irrelevant to a team that deploys Python services, however impressive it looks.

Per language views such as Rust, TypeScript, and Python make this quick. Pick one, then scan the top twenty.

Read the README like a buyer

For each candidate, spend two minutes on the README and answer four questions:

1. Can you say what it does in one sentence? If the README cannot, the maintainers may not know yet either.
2. Is there a quick start that runs? A copy and paste install that works is a strong sign of care.
3. Does it name its limits? Mature projects say what they do not do. Hype projects claim everything.
4. Who is behind it? A known maintainer, a company, or a single weekend project. None of these are bad, but they set different expectations.

Check the activity underneath the stars

Stars are cheap. Work is not. Open the repository and look at:

• Recent commits from more than one person
• Issues opened by people outside the team, and whether anyone answers them
• Releases or tags, even early ones, which show the team thinks about versions
• A license file, without which you cannot safely use the code at all

A project with steady commits, answered issues, and a clear license is far more likely to still exist in six months than one with a viral launch and a silent issue tracker.

Keep a short watchlist

Do not try to follow everything. Keep a list of five to ten projects and revisit it every couple of weeks. Drop the ones that stall, promote the ones that keep shipping.

When one of them announces a beta or a launch date, that is the moment to try it for real. If it is a tool other developers should know about, submit it to TeaserTrack so its launch shows up on the calendar.

A ten minute weekly routine

1. Open Rising on GitHub filtered to your main language and the past week.
2. Skim the top twenty descriptions and open anything relevant.
3. Apply the README questions and the activity check.
4. Add at most two projects to your watchlist.
5. Remove anything that has gone quiet.

Ten minutes a week is enough to stay ahead of what your team will be talking about next quarter.