What This Is
This is the actual automation pipeline behind my daily LinkedIn content. Instead of staring at a blank page every morning, it does the heavy lifting: it researches trending AI & startup topics, writes posts in a defined voice, designs branded carousels and data infographics, and can deliver them to Slack and help you schedule them.
It's not a polished SaaS — it's the real working toolkit, shared as-is so you can run it, learn from it, and bend it to your own brand.
What It Does
AI Post Writing
Generates ready-to-post LinkedIn text via OpenRouter, Anthropic or Gemini.
Branded Carousels
Builds multi-slide carousels as PNG + PDF from HTML templates.
Data Infographics
Renders clean single-image infographics for performance posts.
Topic Research
Pulls trending topics from Reddit & AI news (Apify / Scrapingdog / RSS).
Slack Delivery
Sends finished posts & assets straight to a Slack channel.
Scheduling Helpers
Scripts to queue and organise posts across the week.
What's Inside the ZIP
- Post writers (multi-model)
- Carousel builders
- Infographic renderers
- Daily paper generator
- Reddit + AI-news fetchers
- RSS fallback
- Slack delivery scripts
- Scheduling helpers
- README + help.txt
- 4 reusable skills
- .env.example template
- sample-outputs folder
What You'll Need
Everything runs on your own accounts and machine — so you control usage and cost.
- Python 3 and Node.js installed.
- A writing model key — OpenRouter, Anthropic, or Gemini (all have free / low-cost tiers).
- Research keys (optional) — Apify or Scrapingdog; there's an RSS fallback if you skip them.
- Slack bot token (optional) — only if you want auto-delivery to Slack.
This download was scrubbed of all secrets — there are no API keys, tokens, or .env file inside. You add your own keys to the included .env.example (rename it to .env).
Quick Start
- Unzip the folder anywhere on your computer.
- Rename .env.example → .env and paste in your own API keys.
- Install dependencies: Python packages (e.g. requests) and run npm install inside carousel-routine/.
- Run a generator — for example python generate_all_content_gemini.py.
- Open README.md and help.txt for the full command list.