Skip to content

DmytroG-dot-com/linkedscraper

Repository files navigation

LinkedIn Feed Watcher

Local Playwright CLI that opens your own LinkedIn feed, scrolls posts, extracts post text + URL, and sends each post to your webhook.

Typical flow:

LinkedIn feed -> webhook -> n8n -> AI fit check -> Google Sheet / Telegram

Important

  • runs locally on your machine
  • uses your own browser session
  • you log into LinkedIn yourself
  • no LinkedIn password goes into config
  • not a cloud service, extension, or phishing flow

What gets sent

The scraper sends only post data:

  • post text
  • post URL
  • author
  • published label

Your portfolio / skills / "knowledge" are not sent by the scraper.

Put that in the n8n template node:

  • Set portfolio + skills knowledge

Template file:

Setup

  1. Install deps
npm install
PLAYWRIGHT_BROWSERS_PATH=0 npx playwright install chromium
  1. Create local config
  • copy config/app.example.json -> config/app.json
  • add your own webhook URLs
  1. Import the n8n template
  • set your webhook path
  • fill Set portfolio + skills knowledge
  • connect OpenAI / Google Sheets / Telegram credentials
  1. Log into LinkedIn
npm run login

That opens a local persistent browser. Log in manually.

Commands

npm run help
npm run login
npm run inspect-feed
npm run scan
npm run scan:test
npm run scan:prod

Config

Public example:

Local runtime config:

  • config/app.json (gitignored)

Common knobs:

  • scan.sendAllVisiblePosts
  • filters.containsAny
  • delivery.target
  • delivery.mode

Notes

  • npm run inspect-feed shows what the parser sees before you send real traffic
  • exact post permalinks are not always available; sometimes LinkedIn only exposes profile/company fallback URLs
  • this is feed-based, not full-profile crawling

About

Will read posts in your Linkedin feed endlessly and send them for AI processing (costs $0.7 for 1000's posts), I use it to evaluate all posts to see if its a job post, then AI compares with my skillset and sends into my telegram if its a good fit. Can be configured for any filtring rule. hi@dmytro.biz if need help.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors