What GStars.dev is
GStars.dev is an index of GitHub repositories ranked by recent momentum — star growth over the last 7, 30 and 90 days — and by practical adoption signals such as release cadence, contributor activity and issue handling. Repositories are organized by use-case category (AI agents, DevOps, databases, security and so on) and by maintainer type: company-backed, indie, or foundation.
It exists because a raw star count rewards age, not current relevance. A project that picked up 40,000 stars in 2016 and has been quiet since outranks one that gained 3,000 stars this month on every default GitHub view. GStars.dev inverts that: the question is what developers are adopting now.
What is original here
Everything on the site that is not a direct GitHub field is computed or written here:
- Growth windows for every tracked repository: stars gained over 7, 30 and 90 days, plus percentage growth and an average daily rate.
- Global rankings and per-category rankings for each window.
- A momentum score that blends absolute and relative growth, and a health score (0–100) built from activity, community and maintenance signals.
- Week-over-week movers: which repositories climbed or fell the most in the 7-day ranking since last week.
- AI-written “Use Cases & Benefits” summaries on each repository page.
- Star-history charts, embeddable as images in READMEs and posts.
Repository pages quote only a short excerpt of each project’s README, clearly attributed and linked back to the source on GitHub. READMEs remain © their authors and are distributed under the license each project chose; GStars.dev does not republish them in full. Descriptions, topics and metadata come from the public GitHub API.
Where the data comes from
Data is pulled from the GitHub REST API and from GitHub’s trending pages. Snapshots of stars, forks, watchers and open issues are collected daily for trending and partner repositories, and in rolling nightly batches (stalest first) for everything else tracked; rankings are recomputed after each collection. The sitemap and the machine-readable llms.txt are regenerated daily. The exact formulas behind the growth windows, momentum score, rankings and health score are documented on the methodology page.
How AI is used
The “Use Cases & Benefits” bullets on repository pages are generated by a language model from the repository’s README, its GitHub description, topics, primary language, license, and the names of its recent releases. Star, fork and growth numbers are deliberately withheld from that prompt, so the summary describes what a project does and who it is for — not how popular it is. Summaries are regenerated when missing and refreshed once they are older than 14 days.
Category assignment is rule-based, not AI: each repository’s GitHub topics are mapped to a fixed, hand-maintained taxonomy; if no topic matches, a keyword match on the description is used, then the primary language. Categories are assigned when a repository is first tracked. The “Use Cases & Benefits” bullets are the only AI-generated text on the site; the numbers, rankings and charts are computed directly from the collected snapshots.
Who runs it
GStars.dev is built and maintained by Jeronimo De Leon, an independent developer. It is not affiliated with GitHub. The footer lists other sites by the same maintainer, and Backblaze, a partner whose developer program is promoted on this site; repositories from Backblaze’s open-source organizations are always included in the index (see Methodology).
Contact & corrections
To request a correction, ask for a repository to be removed from the index, or report a problem with the site, email [email protected], or reach out on X at @gstarsdev. Removal requests from a repository’s owner are honored. For how the site handles analytics, advertising cookies and account data, see the privacy policy.