Head over to Google BigQuery and their collection of GitHub Timeline data.
https://bigquery.cloud.google.com/table/githubarchive:github.timeline
Find the total number of created repositories that are public and known to Google BigQuery.
SELECT COUNT(repository_url) FROM [githubarchive:github.timeline] WHERE type = "CreateEvent" AND payload_ref IS NULL
Select a subset of the repositories by URL. This will find roughly 10,000 repository URLs.
SELECT repository_url FROM [githubarchive:github.timeline] WHERE type = "CreateEvent" AND payload_ref IS NULL AND RAND() < 10000/13626213
Export the results to CSV.
Then, pull down this repository.
git@github.com:kdaigle/random-repositories.git
Run bundle install. Set the path_to_bigquery_csv variables to the full file path to your CSV. I've included a sample file here that you can replace.
Setup a GitHub personal access token with public_repo scope.
Run the script with your personal access token. Note: you can only make 5,000 requests per hour to the API. If you need more repositories than that, you'll need to adjust the script to split into groups.
ACCESS_TOKEN=yourPersonalAccessToken ruby random_repos.rb