GitHub Scraper: Repos, Releases, Issues, PRs & Contributors

Export GitHub data to JSON, CSV or Excel without writing a line of API code. Everything comes from GitHub's official REST API, so nothing here breaks when GitHub redesigns its website.

One Actor, seven jobs:

Mode	What you get
Repository details	Stars, forks, watchers, open issues, language, licence, topics, size, default branch, archived and fork flags, created/updated/pushed dates
Releases	Every release with its tag, notes, author, dates and per-asset download counts, plus a total across assets
Issues	Title, state, body, author, labels, assignees, milestone, comment count and dates
Pull requests	The same plus merged state, draft flag, head and base branches and requested reviewers
Contributors	Every contributor with their commit count
Commits	SHA, subject, full message, author and date (author email on request), optionally filtered by branch and start date
Repository search	Any GitHub search query, ranked, with full repository stats for each hit

Why use this one

• Release download counts. Most GitHub scrapers stop at stars. This one totals the download count of every binary attached to every release, which is the only public measure of how much a desktop or CLI tool is actually used.
• Seven datasets, one Actor. Track a competitor's repository, its release cadence, its issue backlog and who is contributing, without stitching together four tools.
• Honest about rate limits. The Actor reads GitHub's rate-limit headers on every response and warns you in the log before the budget runs out, instead of silently returning half a dataset.
• Pull requests are separated from issues. GitHub's issues endpoint quietly mixes pull requests in. This Actor filters them out by default so "Issues" means issues.

Read this before your first run: the rate limit

GitHub allows 60 requests an hour from an anonymous IP address. That is enough for about 60 repositories in Repository details mode, or 6,000 issues, and nothing more.

Adding a free personal access token raises the limit to 5,000 requests an hour. The token needs no scopes at all for public data.

1. Go to github.com/settings/tokens and generate a fine-grained or classic token.
2. Tick no scopes. Public data needs none.
3. Paste it into the GitHub token field. It is stored encrypted and never appears in the dataset.

Repository search is limited separately: 10 queries a minute anonymously, 30 with a token.

How to use it

1. Choose What to scrape.
2. Add Repositories as owner/name or GitHub links, or write Search queries.
3. Paste a GitHub token.
4. Click Start, then download the results or pull them through the Apify API.

Example input: every release of a repository with download counts

{
  "mode": "releases",
  "repositories": ["yt-dlp/yt-dlp"],
  "maxResultsPerRepo": 50,
  "githubToken": "ghp_your_token_here"
}

Example input: open bugs across several repositories

{
  "mode": "issues",
  "repositories": ["apify/crawlee", "scrapy/scrapy"],
  "issueState": "open",
  "labels": ["bug"],
  "maxResultsPerRepo": 500
}

Example input: find fast-growing Python projects

{
  "mode": "searchRepos",
  "searchQueries": ["language:python stars:>5000 pushed:>2026-01-01"],
  "searchSort": "stars",
  "maxResultsPerRepo": 100
}

Output

Each item carries a type field: repo, release, issue, pullRequest, contributor or commit.

Repository

{
  "type": "repo",
  "repo": "apify/crawlee",
  "owner": "apify",
  "description": "Crawlee—A web scraping and browser automation library",
  "language": "TypeScript",
  "stars": 15234,
  "forks": 661,
  "watchers": 92,
  "openIssues": 118,
  "license": "Apache-2.0",
  "topics": ["crawler", "scraping", "playwright"],
  "defaultBranch": "master",
  "isArchived": false,
  "pushedAt": "2026-10-15T09:21:44Z",
  "url": "https://github.com/apify/crawlee"
}

Release

{
  "type": "release",
  "repo": "yt-dlp/yt-dlp",
  "tag": "2026.10.01",
  "name": "yt-dlp 2026.10.01",
  "publishedAt": "2026-10-01T22:14:03Z",
  "isPrerelease": false,
  "assetCount": 12,
  "totalDownloads": 2841905,
  "assets": [
    { "name": "yt-dlp.exe", "downloads": 1204551, "sizeBytes": 12894720, "contentType": "application/octet-stream" }
  ]
}

Pricing

Pay only for results:

Result	Price
Repository (details or search hit)	$5.00 per 1,000
Release, issue, pull request, contributor or commit	$1.00 per 1,000

Set a maximum spend per run in the run options and the Actor stops cleanly when it is reached.

Limits to know about

• The rate limit is the real constraint, not this Actor. See the section above. Without a token, expect runs to stop early.
• Repository search returns at most 1,000 results per query, a GitHub limit. Narrow the query with stars:, pushed: or language: ranges to go deeper.
• Contributors are capped at 500 by GitHub on very large repositories, and anonymous contributors are not included.
• watchers is GitHub's subscriber count. GitHub's own watchers_count field is a duplicate of the star count, so it is deliberately not used here.
• Private repositories need a token with repository access. Without one they return nothing.

FAQ

Do I need a token? No, but a run without one is limited to 60 requests an hour, which is very little. A free token with no scopes lifts it to 5,000.

Is my token safe? It is stored as an encrypted secret input, sent only to api.github.com, and never written to the dataset or the log.

Can I track a repository daily? Yes. Create a schedule in Apify Console and connect the dataset to Slack, Google Sheets, Zapier or Make.

Why are some issues missing? By default pull requests are filtered out of Issues mode. Turn on Include pull requests in Issues mode to keep them.

Feedback

Found a bug or need a field that is missing? Open an issue on the Actor's Issues tab.