How it works
A cultural machine with receipts.
It pulls public signals into a readable field, then keeps the method visible: source, score, mood, confidence, and what still needs a human eye.
The operating manual for the pull
01
Pull signals
The prototype gathers sample cultural signals such as new releases, ratings, review buzz, and availability links.
02
Merge duplicates
Similar titles and different source signals are treated as one cultural item, so the feed never repeats itself.
03
Summarize mood
AI-style summaries turn scattered signals into a readable cultural mood — short, original, and clearly labeled.
04
Publish to Framer CMS
Every item becomes an editable CMS entry that can power pages, cards, filters, and detail views.
Latest prototype crawl
September Cultural Pull #14
Found
24
Created
9
Updated
9
Merged
3
Prototype crawl. Merged three duplicate editions and normalized two re-release titles. CMS provided the curated sample data.
Source ethics & transparency
This prototype avoids heavy scraping and copyrighted review reuse. It uses CMS seed data, original summaries, and source-style links to demonstrate how an external agent could later update the feed.
No live crawling happens in this version. Scores and momentum are illustrative sample values, summaries are original and clearly labeled as AI-style, and every source or availability link is a placeholder you can replace in the CMS.
Enter the latest pull