GitHub Repository Insights

Turn GitHub repositories into clean, structured data. Point this actor at specific repos, at whole users/organizations, or at a GitHub search query, and get back one tidy JSON row per repository — stars, forks, watchers, open issues, language, topics, license, homepage and timestamps.

It runs on the official GitHub REST API. No login and no proxy are needed; an optional personal access token only raises the rate limit.

Why use it

• Three ways in, one shape out. Mix explicit repos, user/org listings and a search query in a single run — every source produces the same flat record.
• Never fails on one bad input. A missing repo (404) or a rate-limit hit (403) becomes a labelled error row; the run keeps going.
• No dependencies to fear. Just the Apify SDK and Node 20's built-in fetch.
• Pay only for what you get. Billed per delivered repository (pay-per-event) — error rows are never charged.

Input

Field	Type	Default	Description
repos	array	—	Repositories to fetch, each as "owner/name" (e.g. "apify/crawlee").
searchQuery	string	—	Optional GitHub repository search query. Returns the most-starred matches first.
users	array	—	Optional list of GitHub usernames/organizations. Fetches each account's repositories.
maxResults	integer	30	Max repositories per source (per user and for the search query). Capped at 100 by the API.
githubToken	string (secret)	—	Optional personal access token. Raises the rate limit from 60 to 5,000 requests/hour.

Provide at least one of repos, users or searchQuery. You can combine all three.

Example input

{
  "repos": ["apify/crawlee", "facebook/react"],
  "searchQuery": "web scraping language:python stars:>1000",
  "users": ["apify"],
  "maxResults": 30
}

Output

Each dataset record is a flat repository object:

{
  "fullName": "apify/crawlee",
  "owner": "apify",
  "name": "crawlee",
  "description": "Crawlee—A web scraping and browser automation library…",
  "stars": 18500,
  "forks": 810,
  "watchers": 18500,
  "openIssues": 130,
  "language": "TypeScript",
  "topics": ["crawler", "scraper", "playwright", "puppeteer"],
  "license": "Apache-2.0",
  "homepage": "https://crawlee.dev",
  "url": "https://github.com/apify/crawlee",
  "createdAt": "2021-05-31T14:15:11Z",
  "updatedAt": "2024-06-01T00:00:00Z",
  "pushedAt": "2024-06-01T00:00:00Z",
  "size": 45000,
  "defaultBranch": "master",
  "archived": false,
  "isFork": false
}

Field	Description
fullName	Repository full name (owner/name).
owner	Owner login.
name	Repository name.
description	Repository description.
stars	Stargazer count.
forks	Fork count.
watchers	Watcher count.
openIssues	Open issues + pull requests count.
language	Primary language.
topics	Array of repository topics.
license	SPDX license identifier (e.g. MIT, Apache-2.0).
homepage	Project homepage URL, if set.
url	Link to the repository on GitHub.
createdAt	Creation timestamp (ISO 8601).
updatedAt	Last metadata update timestamp (ISO 8601).
pushedAt	Last push timestamp (ISO 8601).
size	Repository size in kilobytes.
defaultBranch	Default branch name.
archived	Whether the repository is archived.
isFork	Whether the repository is a fork.

When an input can't be fetched, its row is { "input": "<value>", "error": "<message>" } instead — the run continues.

Rate limits

The GitHub REST API allows 60 requests/hour unauthenticated. Providing a githubToken raises this to 5,000 requests/hour; a read-only token with no scopes is enough. Hitting the limit produces an error row (HTTP 403) rather than crashing the run.

Data source

This actor uses the official GitHub REST API:

• Base URL: https://api.github.com
• Repo details: GET /repos/{owner}/{repo}
• User repos: GET /users/{user}/repos?sort=updated
• Search: GET /search/repositories?q=<query>&sort=stars&order=desc

Requests send the required User-Agent and Accept: application/vnd.github+json headers.

Billing

This actor uses the pay-per-event pricing model. You are charged once per repository successfully delivered to the dataset (event repo-scraped). Error rows are never charged.

---

Built by Assia Fadli. Licensed under MIT.