How do you promote a GitHub project without feeling like a spammer?

Here is the thing nobody tells you when you ship: the reason self-promo feels gross is that most of it actually is gross. Not you. The format. Dropping a bare link into a thread that was not about you, with a "Check out my project!" and a rocket emoji, is not promotion. It is litter. Your body knows it, which is why you feel the ick before you even hit post.

The reframe that fixed this (and it is small, and it is annoying, and it works): stop trying to promote the project, and try to be useful in public instead. Same energy, opposite feeling. When you answer someone's actual question, or write down the thing you learned the hard way, the link to acme/rocketdb becomes a footnote to something helpful, not the whole point of the message. Nobody has ever felt spammed by a person who was genuinely trying to help them.

You do not have to become an extrovert. (I would know.) You just have to move the value to the front and the ask to the back, or leave the ask off entirely and trust people to find it.

The instinct is to post where it is easiest, which is usually your own timeline, to the fourteen people who already like you. That feels safe. It also does close to nothing. Promotion is not broadcasting into your own room. It is going to the room where the people with your problem already gather.

So do the boring homework first. Who is acme/rocketdb for? Where do those people complain about the thing you fixed? A specific subreddit, a Discord, a language-specific forum, the comments under a related post, a newsletter that already covers your niche. That is where you show up, and not to plug. Answer three questions before you ever mention you built something. Earn the room before you use it.

If you are staring at a blank list of rooms, there is a whole guide on finding your first users for an open-source project. The short version: they are almost never on the platform that is most convenient for you.

People do not search for your product. They cannot. They do not know it exists yet. They search for their problem. So write about the problem.

"How I stopped X from happening" will always outperform "Introducing acme/rocketdb v2." The first is a story someone can see themselves inside. The second is a press release, and nobody has ever forwarded a press release to a friend. Write the debugging session, the wrong turn you took, the config that finally worked. Put the tool in as the thing that helped, not the hero. (The trick, which is not really a trick: the honest post about the problem is also just easier to write, because you actually lived it.)

This is slower. It does not spike. But a good problem-post keeps getting found months later by someone typing your exact pain into a search bar, and that person shows up already half-trusting you.

Launch day is a sugar high. You post to the big aggregators, the number goes up, your notifications light up, and for about seventy-two hours you feel like a genius. Then it fades. It always fades. (You can watch the exact shape of it on a star-history chart: the vertical cliff of launch day, then the long flat plain after.) This is normal. It is not failure. It is physics.

The mistake is treating launch as the strategy. Launch is one day; momentum is a habit. The projects that compound are not the ones with the biggest spike, they are the ones that keep posting quiet proof: a changelog, a fix, a small before-and-after, a user's win you got permission to share. None of it is loud. All of it says "this thing is alive and getting better," which is the only signal that turns a curious star into a real user.

So after the confetti, keep turning that interest into something real with the least glamorous work there is: showing up next week, and the week after, with one more small true thing to say.

Here is the channel everyone skips, and it is the best one: the developers who already starred your repo. They raised their hand. They saw acme/rocketdb, felt something, and clicked the one button GitHub gives them to say so. That is the warmest audience you will ever have, and most people ignore it entirely to go chase colder strangers.

The catch (there is always a catch) is that GitHub restricted the public stargazers list in 2026 to a repo's own admins and collaborators, so the general public can no longer browse who starred you. As the owner you keep that access, but even a full page of usernames is not the same as knowing who those people are. The count is still public. The people got harder to know. But a star count is aggregate: it tells you how many, not who. And the who is the entire value. dev-priya starring you is worth more than a hundred anonymous ticks, because dev-priya is a person you could actually talk to.

This is the one honest place a tool earns its keep. WarmStars takes your OWN repo's stargazers and turns them back into named people from public data, with a public email for many of them, so you can send a real, specific thank-you instead of shouting into a feed. Not a blast. A note: you starred acme/rocketdb, I am the person who built it, was it the thing you needed? See how to see who starred your repo and how to actually reach them. That is the least cringe promotion there is: talking to someone who already told you they were interested.