This Open-Source Tool Solved Scraping The Web (Scrapling)
Join AI Made Simple (free): https://www.skool.com/ai-automation-n...
Work with DeepRooted AI: https://deeprooted.io
There is a free Python library called Scrapling that just passed 71,000 stars on GitHub, and it lets your AI agent read a web page for a fraction of what you are paying right now. In this one I run it through three real workflows on camera: I cut the same page from 215,000 tokens down to about 3,000, I turn a public Better Business Bureau directory into a qualified prospect list, and I scrape five med spa sites and build my own landing page out of what they all had in common. I also show you where it falls over, because the claim going around that it bypasses every bot protection on the internet is not true, and independent testing puts Cloudflare success closer to 58 percent.
🛠 My Tools:
Glaido🎙 :
https://get.glaido.com/kacper-rutkiewicz
🔗 Resources
Scrapling on GitHub:
https://github.com/D4Vinci/Scrapling
Scrapling docs:
https://scrapling.readthedocs.io
Imperva Bad Bot Report 2026:
https://www.imperva.com/resources/res...
Firecrawl (the hosted API I compared against):
https://www.firecrawl.dev
TIMESTAMPS:
00:00 - Intro
00:45 - What web scraping actually is
01:45 - Four reasons anyone scrapes
02:45 - Workflow 1: cutting the token bill in Claude Code
04:35 - The same page through Firecrawl
05:25 - 215,000 tokens down to 3,000
06:15 - Workflow 2: a directory into a prospect list
07:25 - Fetcher, StealthyFetcher, DynamicFetcher
08:35 - Letting Claude qualify the leads
10:00 - Workflow 3: five sites, one landing page
13:00 - The finished page
14:00 - So what Scrapling actually is
15:55 - The legal line and the Cloudflare claim
16:35 - How to install it
17:45 - Outro
18:25 - Blessing
Connect on LinkedIn💼 / kacper-rutkiewicz
#scrapling #claudecode #webscraping