twitter-scraping - SKILL.md Agent Skill

name: twitter-scraping description: Scrapes X/Twitter data without API access using Puppeteer stealth and browser console scripts. Extracts profiles, followers, following lists, tweets, search results, hashtags, threads, media, bookmarks, notifications, DMs, likes, and viral tweets. Exports to JSON/CSV. Use when collecting, exporting, or analyzing Twitter data. license: MIT metadata: author: nichxbt version: "4.0"

Twitter Scraping

Auth Setup

All scrapers require an auth_token cookie from x.com (DevTools -> Application -> Cookies -> copy auth_token value).

Node.js Scraper API

npm install xactions -- all functions from src/scrapers/index.js.

import { createBrowser, createPage, loginWithCookie, scrapeProfile } from 'xactions';

const browser = await createBrowser();
const page = await createPage(browser);
await loginWithCookie(page, AUTH_TOKEN);

const profile = await scrapeProfile(page, 'nichxbt');
await browser.close();

Function	Purpose
`scrapeProfile(page, username)`	Profile data (bio, followers, following)
`scrapeFollowers(page, username, {limit})`	Follower list
`scrapeFollowing(page, username, {limit})`	Following list
`scrapeTweets(page, username, {limit})`	User's tweets
`searchTweets(page, query, {limit})`	Search results
`scrapeHashtag(page, tag, {limit})`	Hashtag tweets
`scrapeThread(page, tweetUrl)`	Thread tweets
`scrapeMedia(page, username, {limit})`	Media URLs
`exportToJSON(data, path)` / `exportToCSV(data, path)`	File export

Browser Console Scripts

Standalone IIFEs -- paste into DevTools console on x.com. No dependencies.

Script	File	Navigate to
Download video	`src/scrapers/videoDownloader.js`	Tweet with video
Export bookmarks	`src/scrapers/bookmarkExporter.js`	`x.com/i/bookmarks`
Unroll thread	`src/scrapers/threadUnroller.js`	Any thread
Find viral tweets	`src/scrapers/viralTweets.js`	User profile
Scrape followers	`scripts/scrapeFollowers.js`	`x.com/USER/followers`
Scrape following	`scripts/scrapeFollowing.js`	`x.com/USER/following`
Scrape profile	`scripts/scrapeProfile.js`	User profile
Scrape search	`scripts/scrapeSearch.js`	Search results
Scrape hashtag	`scripts/scrapeHashtag.js`	`x.com/hashtag/TAG`
Scrape likes	`scripts/scrapeLikes.js`	`x.com/USER/likes`
Scrape likers	`scripts/scrapeLikers.js`	Any tweet
Scrape replies	`scripts/scrapeReplies.js`	Any tweet
Scrape media	`scripts/scrapeMedia.js`	User media tab
Scrape DMs	`scripts/scrapeDMs.js`	`x.com/messages`
Scrape notifications	`scripts/scrapeNotifications.js`	`x.com/notifications`
Scrape list	`scripts/scrapeList.js`	Any X List
Scrape bookmarks	`scripts/scrapeBookmarks.js`	`x.com/i/bookmarks`
Export to CSV	`scripts/exportToCSV.js`	After any scrape
Link scraper	`scripts/linkScraper.js`	Any timeline
Quote retweets	`scripts/scrapeQuoteRetweets.js`	Any tweet

MCP Scraping Tools

All scrapers available via the MCP server for AI agents:

Tool	Description
`x_get_profile`	Scrape user profile
`x_get_followers`	Get follower list
`x_get_following`	Get following list
`x_get_tweets`	Get user tweets
`x_search_tweets`	Search tweets
`x_get_thread`	Unroll thread
`x_download_video`	Extract video URL

Notes

Always call browser.close() when done with Node.js scrapers
Browser scripts stop on page navigation -- stay on the page while running
All scripts support JSON and CSV export
Increase limit parameter for deeper scraping (default varies by function)
Rate limiting is built in -- avoid reducing delays below defaults