Frequently asked questions
Short, factual answers about GitHub repository traffic and this dashboard.
How long does GitHub keep repository traffic data?
GitHub keeps only the last 14 days of views and clones, and only the top 10 referrers and popular paths. Older days are discarded, so this dashboard stores a snapshot every day to keep a permanent history.
How often is the dashboard updated?
A scheduled GitHub Actions workflow runs once a day. GitHub can take 24-48 hours to finalise a day's numbers, so the most recent days may still increase.
Does it show private repositories?
No. Private repositories are never read, stored or published. Only public repositories appear here.
Why are clones much higher than views?
Clones include automated traffic such as CI systems, mirrors and crawlers, so they are a weaker signal of human interest than views. Views and unique visitors are the better indicators.
What are unique visitors and why are they not added up over months?
GitHub reports a deduplicated unique-visitor estimate for its 14-day window only. Adding daily uniques would count the same person repeatedly, so longer periods show views and clones, not summed uniques.
How is the repository score calculated?
score = views in 30 days + 0.5 x clones in 30 days + 2 x unique visitors (14 days) + 3 x stars + 4 x forks. The top 20% of public, non-fork, non-archived repositories are marked top and the bottom 20% low.
How are API rate limits avoided?
Repositories with recent views are fetched daily, others every 3 days and dormant ones weekly. Because GitHub returns a 14-day window on every call, no data is lost. Calls are sequential, and the run stops safely if the remaining hourly budget gets low.
Can I run this for my own GitHub account?
Yes. It is open source under the MIT license: fork github.com/musman550/github-traffic-analytics-dashboard, add a GH_PAT secret with the repo scope, enable GitHub Pages on the main branch /docs folder and run the workflow.
Who can change this dashboard?
Only the repository owner can change the code and data. The main branch is protected, forks cannot write to it, and the site runs under a strict content-security policy.
How many repositories are tracked right now?
276 public repositories, with data through 2026-11-17.
Can I see who visited my GitHub repository?
No. GitHub's traffic data is aggregate only: counts of views and clones, unique-visitor estimates and referrer sites. It does not identify individual visitors.
Who can see repository traffic on GitHub?
Only people with push access to a repository can read its traffic data, in the Insights tab or through the REST API.
Why does github.com appear as a referrer?
It usually means visitors arrived from another page on GitHub, such as a profile, search results or a different repository.
How is this different from GitHub's Insights tab?
Insights shows only the last 14 days for one repository at a time. This dashboard stores every day, combines all your public repositories and adds trends, sources, rankings and exports.
How do I export the data?
Download repos.csv and daily.csv, or the JSON files, from the data page. They are static files that update daily.
What happens if an update fails?
The status page shows every collection run. A failed or partial run creates an alert issue, and a daily watchdog checks that the published data is fresh. Because each GitHub call returns 14 days, a missed day is recovered by the next run.
Can I install the dashboard as an app?
Yes. Browsers that support installable web apps offer an install option, and the last loaded data stays available offline.
Does AI change my repositories?
An optimizer proposes better descriptions, topics and a README summary for the owner's public tool repositories. Topics are only added, README changes arrive as pull requests, every change is logged on the optimizer page and a traffic drop triggers a rollback.
Is this affiliated with GitHub?
No. It is an independent open-source project that uses the public GitHub REST API.