Competitor Ads Scraper (Claude Code Template)

Pull your competitors' ads from the Meta Ad Library straight into Airtable — ad copy, headlines, creatives (images + videos), targeting, platforms, and more.

Want to go deeper?

This template is just one piece of the puzzle. Inside my Mr Paid Social community on Skool you'll get:

Hands-on help with this scraper (and every other AI ad tool I build)
A growing library of AI workflows for creative, copy, and media buying
Live trainings and a community of media buyers leveling up their Meta ads with AI

Join us at skool.com/mrpaidsocial →

Getting Started

1. Open Claude Code

Open Claude Code (CLI, desktop app, or IDE extension).

2. Paste this repo URL and say "set this up"

That's it. Claude will clone the repo, install dependencies, walk you through API key setup, and get everything running.

What you'll need

Python 3.10+ — python.org/downloads
Google Chrome — needed for extracting ad creatives (runs headless, you won't see a browser window)
Meta access token — free, takes 30 seconds: developers.facebook.com/tools/explorer
Airtable account + Personal Access Token — free tier works: airtable.com/create/tokens

What you get

A Competitor Ads Airtable table with 20 fields per ad:

Field	Description
Competitor Ad Name	Auto-generated descriptive name
Facebook Page	Page running the ad
Creative	Image or video attachment (auto-downloaded)
Ad Copy	Full ad text
Headline / Description / CTA	Link preview fields
Landing Page URL	Where the ad sends traffic
Ad Library URL	Direct link to view in Meta's Ad Library
Start Date / Last Seen Date	When the ad ran
Active Status	Active or Inactive
Platforms	Facebook, Instagram, Messenger, Audience Network, Threads
Impressions	High / Medium / Low (derived from EU reach data)
Target Ages / Locations	Audience targeting
...and more	Languages, EU Total Reach, Page ID, Ad Library ID

Two ways to find competitors

Direct — give Claude a list of Facebook page names or IDs
Discover — give Claude a URL (like your website or a competitor's Skool page) and it will suggest relevant competitors automatically

How it works under the hood

pull_ads.py calls the Meta Ad Library API (ads_archive) for each competitor
Parallel headless Chrome workers extract creative images/videos from ad snapshots
pyairtable batch-inserts all records directly to your Airtable base (~20 seconds for 100+ ads)
A JSON backup is saved locally

See CLAUDE.md for the full Claude Code runtime flow and META_ADS_LIBRARY_API.md for the API reference.

SMontiel/Meta-Ads-Spy-Claude-Code-Airtable

README