Before we begin, let me do a quick self-promotion:
My new webapp is live here (If you’re struggling with what to build, I got you)
Remember, when GitHub stars were treated like the ultimate technical currency?
Oh maybe they still are!
Well, the reality is:
If a metric can be gamed, it will be.
When I was pushing the initial open-source releases for my projects in college, I used to watch the repository obsessively.
Earning an organic GitHub star was brutally difficult.
It required writing solid code, drafting clear documentation, and solving a real problem for another human being.
While we are treating GitHub stars as a proxy for product-market fit, a lot of startups are treating them as a cheap marketing commodity to scam Venture Capitalists.
Let’s start with:
The Fiverr Black Market
A brutally simple economic reality.
You can literally buy GitHub stars on Fiverr right now for $0.45 per star.
This is not a niche operation.
It is a highly professionalized shadow economy.
Startups are bypassing the brutal grind of community building and just purchasing bulk engagement.
Why?
Because Venture Capital firms use GitHub stars as a primary metric to evaluate traction for seed rounds.
If spending $4,500 on Fiverr buys you 10,000 stars, and those 10,000 stars convince a VC to write a $2,000,000 seed check, the unit economics of the fraud are irresistible.
The Carnegie Mellon Teardown
Carnegie Mellon University recently published a bombshell study utilizing a detection tool called StarScout.
The researchers analyzed millions of accounts and exposed 6 million suspected fake stars distributed across over 18,000 repositories.
Brb, crying?
They found two primary patterns of manipulation:
- Low-Activity Throwaways: Disposable accounts created specifically to click the star button and never commit code.
- Clustered Amplification: Massive, coordinated starring campaigns that hit repositories in tight timeframes to force the project onto the “GitHub Trending” page.
Once a project hits the Trending page, real developers download it, creating a layer of organic camouflage over the artificial inflation.
The AI Startup Exploitation
The most alarming part of the original teardown is who is actually buying these stars.
It is not just script kiddies looking for clout.
It is heavily funded startups.
The papers explicitly called out Langflow, a massive open-source project.
As of the report, Langflow is sitting on roughly 148,000 stars.
According to the Carnegie Mellon data, an estimated 48% of those stars are flagged as fake.
AI wrappers and frameworks are the biggest offenders right now.
The hype cycle creates immense pressure to show immediate traction. Instead of building better agentic workflows, founders are purchasing clustered engagement to fake an active community.
So, Stop Trusting Vanity Metrics like stars.
If you are a developer, an investor, or an enterprise decision-maker, you have to stop using the star count as a primary filter for adoption.
A 4,000-star repository might just be a $1,800 marketing budget in disguise.
Look at the fork-to-star ratio.
Check the issue discussions to see if actual humans are reporting bugs.
Audit the commit rhythm.
If the open-source ecosystem does not collectively stop treating a clickable star as a sign of engineering quality, the entire system collapses into a pay-to-play marketplace.
In case we are meeting for the first time, come over here, it’ll be worth the roller coaster of articles that are gonna come up in the next few weeks.
I swear tracking these updates is a job in itself, lately.
Here’s the list which I’ve built and keep adding on.
And If you need help for analyzing UFC fights, please check out BoutPredict :)