Skip to content

AI-Answer/content-factory

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Content Factory

Repo: github.com/AI-Answer/content-factory
Live demo: glacial-geyser-vwbp.here.now/app/

A pre-built content strategy system you run locally.

The value is not the scraper. It is the intelligence, logic, and process — already designed:

Research winners first → understand why they work → choose an angle → produce deliberately → publish as a trial → learn and repeat.

Most people can ask AI for post ideas. What they lack is the operating logic: what to study, how to judge virality, how to extract patterns, when to approve an angle, which format fits, and how performance feeds next week. That thinking is the product. The code makes it repeatable.


What you need

  1. Python 3.10+
  2. An Apify API keyget one here
  3. 3–5 competitor Instagram accounts in your niche

Start

git clone https://github.com/AI-Answer/content-factory.git
cd content-factory
python3 -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
python3 -m factory init
python3 -m factory serve

Open http://localhost:8000 — the cockpit opens on The system until you scout your first competitors.

Demo: glacial-geyser-vwbp.here.now/app/?workspace=empty#system (empty workspace — learn the process first) · with sample data


The workflow (what the dashboard teaches)

Stage What happens Why You decide
① Collect evidence Scrape competitor reels Strategy needs proof, not vibes Which accounts to study
② Study the market Rank breakouts, extract hooks/themes/formats The idea matters more than polish Which patterns fit your audience
③ Choose angles Ideas tied to source reels Generic prompts fail; angles need evidence Approve or skip each idea
④ Make the asset Scripted / remake / avatar / carousel Same angle, different execution paths Script, film, advance status
⑤ Publish & learn Schedule + performance snapshots One number means little; history guides next week Repeat, revise, or retire

Each stage in the dashboard explains: what happens · why it matters · what you decide · what moves forward.


Production modes (from the reference guides)

Mode When to use
Scripted Talking-head authority — record, cut silence, captions, music
Remake Winning format is B-roll + trending audio + on-screen text
Avatar Scale without filming — HeyGen + B-roll + render
Carousel Educational/list content as branded slides

Use Claude Code in this repo for the judgement-heavy steps (editing, rendering, scheduling).


Repo layout

factory/       Python engine (scout, ideate, create, schedule, track)
dashboard/     Strategy cockpit — process + workspace
data/          Your local database (gitignored)
guides/        Private reference (techniques, not shipped as product — see CREDITS.md)

Teaching angle

Share this as a Claude Code community template: learn the full short-form content operating system, run it locally, adapt the logic in factory/, and extend with your own voice and niche.

See CREDITS.md for attribution on adapted techniques.

About

Pre-built content strategy workflow for Claude Code — scout, choose angles, produce, learn.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors