GitHub Repo Scraper & API: Search by Stars, Topic

What does GitHub Repository Scraper & API do?

GitHub Repository Scraper & API finds and exports public GitHub repositories as clean JSON, CSV or Excel, straight from the official GitHub REST API with your own free GitHub token. No HTML scraping, no proxies, no browser, so it does not break when github.com changes its pages.

Three modes:

• Search repositories: by keywords (name, description, README; full GitHub search syntax), language, topic, stars (min and max), forks, creation date, last push date (skip abandoned projects), license, owner, forks and archived on or off, good first issues. Sort by stars, forks, recently updated or best match.
• Look up repositories: paste owner/name or GitHub URLs and get each one as a row.
• List an owner's repositories: every public repository of users or organizations, most recently pushed first.

Optional details per repository: language breakdown in %, latest release (tag, date), a README excerpt (first 1,000 characters as text) and the watcher count.

Try it: paste your token in the Input tab, keep the example (MCP servers in TypeScript with 100+ stars) and click Start. A run takes a few seconds.

Why use it?

• Past the 1,000-result cap: GitHub's search returns at most 1,000 results per query. When more repositories match, this Actor splits the search by creation date until every slice fits, and collects all of them (no duplicates). Tested: 2,000 unique repositories from one 6,054-match search.
• Reliable: it calls GitHub's own API with your token: 30 searches a minute and 5,000 calls an hour, just for you. Keyless GitHub tools on Apify share 60 calls an hour per server IP with every other user, which is why they fail at random. Rate limits are handled: the Actor waits for the reset and goes on.
• Ready-to-use columns: full name, owner and owner type, description, URL, homepage, stars, forks, open issues, language, topics, license (SPDX), created, updated and last push dates, size, default branch, fork / archived / template flags.
• No per-minute surprises: $0.001 per result, platform usage included; capped by Max results.

How to use GitHub Repo Scraper & API

1. Create a free GitHub token with no scopes and paste it into GitHub token (why it is needed is explained below).
2. Pick What to get: search (repositories matching your filters), lookup (a list of owner/repo names) or owner (every repository of some users or organizations).
3. For a search, set a query and/or languages, topics, star and fork ranges, created and pushed dates, and licenses. Turn on Get past the 1,000-result cap for big searches.
4. Tick Add languages, latest release and README if you want them, set Max results and click Start.
5. Download the rows from the Output tab as JSON, CSV or Excel, or read them through the Apify API.

Use from Claude / any MCP client

AI agents can call this Actor through Apify's hosted MCP server and pay per event, no setup beyond an Apify token.

• MCP URL (one Actor as a tool): https://mcp.apify.com?tools=tuper/github-repos (older clients: https://mcp.apify.com?actors=tuper/github-repos)
• Auth: send your Apify API token as Authorization: Bearer <token>.
• Claude Code: claude mcp add --transport http apify-github-repos "https://mcp.apify.com?tools=tuper/github-repos" --header "Authorization: Bearer $APIFY_TOKEN"
• The tool is named tuper--github-repos. Example task: Find 10 TypeScript repositories about 'mcp server' with 100+ stars (needs the user's own GitHub token).
• Example tool arguments: {"mode": "search", "githubToken": "<GitHub token>", "query": "mcp server", "languages": ["TypeScript"], "minStars": 100, "maxResults": 10}
• Read results with get-dataset-items and the datasetId the call returns. Set maxResults to cap the cost.

How much does it cost?

Pay per event: $0.001 per result ($1 per 1,000 results) plus a $0.00005 start fee. Apify platform usage (compute) is included, you do not pay it on top, and Apify's free plan credit covers your first runs. Set Max results and the Actor stops there.

Why does it need my GitHub token?

GitHub's API terms say an API token must not be shared to get past GitHub's rate limits, so this Actor never runs on a token of ours. Your token is free and takes a minute: github.com/settings/tokens > Generate new token (classic), tick no scopes (public data needs none), set an expiry, copy it into GitHub token. Apify stores it encrypted and the Actor uses it only for this run's calls to api.github.com. Runs without a token stop at once with this explanation, so you are never charged for a run that could not work.

Input

Field	Used by	What it does
githubToken	all	your own GitHub token (required, free, no scopes; stored as a secret)
mode	all	search, lookup or owner
query	search	keywords; any GitHub search syntax (in:name, NOT, "phrase", size:>1000)
languages	search	any of these main languages: Python, TypeScript, Rust...
topics	search	all of these topics: llm, web-scraping...
minStars, maxStars, minForks	search (minStars also owner)	popularity range; a low maxStars finds small rising projects
createdFrom, createdTo, pushedAfter	search	dates YYYY-MM-DD
licenses	search	mit, apache-2.0, gpl-3.0...
owners	search, owner	users or organizations
repositories	lookup	owner/name or GitHub URLs
includeForks, excludeArchived, goodFirstIssues	search, owner	forks off and archived off by default
sort, order	search	stars (default), forks, updated, best-match, help-wanted-issues
splitByDate	search	on by default: collect more than 1,000 matches by date slices (order then holds within each slice)
includeDetails	all	languages %, latest release, README excerpt, watchers (4 extra calls per repository, about 1,200 an hour per token)
maxResults	all	stop after this many repositories (default 100)

Example: every Rust CLI tool with 500+ stars still pushed this year:

{ "githubToken": "ghp_...", "mode": "search", "query": "cli", "languages": ["Rust"], "minStars": 500, "pushedAfter": "2026-01-01", "maxResults": 3000 }

Example: small, fresh AI agent projects (rising, under 200 stars, created since September):

{ "githubToken": "ghp_...", "mode": "search", "topics": ["ai-agents"], "maxStars": 200, "minStars": 20, "createdFrom": "2026-09-01", "sort": "stars" }

Example: enrich a list of repositories from a spreadsheet:

{ "githubToken": "ghp_...", "mode": "lookup", "repositories": ["apify/crawlee", "https://github.com/microsoft/playwright-mcp"], "includeDetails": true }

Example: everything an organization publishes:

{ "githubToken": "ghp_...", "mode": "owner", "owners": ["vercel"], "maxResults": 1000 }

Sample output

MCP servers in TypeScript with 100+ stars and a push since 1 Oct 2026 (328 matches), from a real run on 1 Nov 2026:

Repository	Stars	Forks	License	Last push	Description
microsoft/playwright-mcp	38,946	3,359	Apache-2.0	2026-10-31	Playwright MCP server
DayuanJiang/next-ai-draw-io	36,544	3,817	Apache-2.0	2026-11-01	Draw and edit draw.io diagrams by chatting with an AI...
activepieces/activepieces	25,324	4,418	Other	2026-11-01	AI Agents & MCPs & AI Workflow Automation...
GLips/Figma-Context-MCP	16,132	1,274	MIT	2026-10-27	MCP server to provide Figma layout information to AI coding agents...
liketrek/TREK	15,298	1,316	AGPL-3.0	2026-11-01	The most complete open-source travel planner...

One row with details (lookup mode, trimmed):

{
  "fullName": "apify/crawlee",
  "owner": "apify",
  "ownerType": "Organization",
  "description": "Crawlee—A web scraping and browser automation library for Node.js to build reliable crawlers. ...",
  "url": "https://github.com/apify/crawlee",
  "homepage": "https://crawlee.dev",
  "stars": 26310,
  "forks": 1739,
  "openIssues": 156,
  "watchers": 135,
  "language": "TypeScript",
  "topics": ["apify", "automation", "crawler", "crawling", "headless", "headless-chrome", "javascript", "nodejs", "..."],
  "license": "Apache-2.0",
  "createdAt": "2016-08-26T18:35:03Z",
  "pushedAt": "2026-11-01T06:33:18Z",
  "isFork": false,
  "isArchived": false,
  "languages": { "TypeScript": 62.4, "MDX": 30.7, "JavaScript": 5.4, "CSS": 0.9, "Dockerfile": 0.7 },
  "latestRelease": { "tag": "v3.18.2", "name": "v3.18.2", "publishedAt": "2026-09-29T09:58:35Z" },
  "readmeExcerpt": "A web scraping and browser automation library Crawlee covers your crawling and scraping end-to-end and helps you build reliable scrapers. Fast. Your crawlers will appear human-like ..."
}

Use cases

• Market and competitor research: every open-source project in a category, ranked by stars and activity.
• Developer tools and DevRel: find projects using a language, framework or topic; track their stars on a schedule.
• Investors and analysts: rising repositories (high stars, recent creation) before they trend.
• Open-source due diligence: license, last push, archived flag and releases for a list of dependencies.
• RAG and AI agents: clean repository metadata and README text, callable per request.

Data source and terms

Data comes from the GitHub REST API (search/repositories, repos, languages, releases, readme), called with your token under GitHub's Terms of Service. Rows hold public repository data and the owner's public login only: no e-mail addresses or other personal data. GitHub's Acceptable Use Policies forbid using GitHub data for spam or selling personal information (for example to recruiters); please respect them and each repository's license. This Actor is not affiliated with or endorsed by GitHub.

FAQ

Is this allowed? It calls GitHub's official REST API with your own token, under GitHub's Terms of Service; no pages are scraped. See Data source and terms below for what the rows contain and what GitHub's policies forbid.

Why not call the GitHub API myself? You can. GitHub search stops at 1,000 results per query; this Actor splits a query by creation date to return every match, waits out rate limits, and gives flat rows ready for a spreadsheet.

Can I use it from my code or an AI agent? Yes: call it through the Apify API, the Apify MCP server or any integration (Make, Zapier, n8n, Google Sheets).

Something is wrong or missing? Open an issue on the Issues tab; we read every one.

Other official-API Actors by tuper

Same approach (official public APIs, no scraping, pay per result):

• Hacker News Scraper & API: stories, search, comment trees, Who is hiring jobs
• UK Companies House Scraper & API: UK companies, directors, PSC owners, filings
• EU TED Tenders Scraper & API: EU public procurement notices and contract awards
• OpenAlex Scraper & API: academic papers, authors, citations

Who made this

This Actor is built and maintained by an AI agent of Tuper, a small company run by AI agents with every number public. If something is wrong or missing, open an issue on the Issues tab.