How TopGit ranks repositories

GitHub is the source of truth

Every repository's facts — stars, forks, topics, primary language, license, and description — originate from GitHub. Metric values reflect the most recent sync. TopGit adds a discovery, ranking, and original editorial layer on top; it never republishes another directory's content.

The quality gate

Not every repository we track is published. Each is scored on real, verifiable metadata — the substance of its description, its popularity and fork activity, its topics, and how recently it was active. That score maps to a lifecycle status:

• Imported / Draft / Needs review — tracked but not indexed (marked noindex).
• SEO-ready — strong metadata, awaiting an original TopGit review before going live.
• Published — passes the quality bar and has an original review; this is the only status we allow search engines to index.

The score is deliberately blind to long-form content, so a page can never be inflated by copied text. Only genuinely original, TopGit-written content moves a page to published.

How trending is ranked

Our trending pages rank repositories by current stars weighted by recency — newer, well-starred repos rank higher, and the daily/weekly/monthly views tighten the recency emphasis. We are explicit about a limitation: TopGit does not yet track precise day-by-day star velocity, so we do not claim “+X stars today” figures. Real velocity tracking is a planned addition, and we will only show those numbers once the underlying data exists.

What we do not do

• No fabricated ratings or reviews. Stars and forks are reported as GitHub interaction counts, never as a 1–5 rating.
• No invented features. If a repository's documentation is thin, we say so rather than guessing.
• No copied content. We do not republish READMEs verbatim or scrape other directories into our pages.
• No indexing of thin pages. Unreviewed or low-quality pages stay out of our sitemap and carry noindex.