This project demonstrates how a Python service could theoretically automate LinkedIn using ChatGPT Web for content generation, comments without token costs — including Search and Image generation.
The research Project worked like this:
- Parse User Feed
- Get random Post with filter list
- Like and Comment
- Add new friends
- Sleep and Continue
This resulted in a fail of ca. 250 followers and much more engagement in the course fo 30 days.
For this to work the Browser must be prepped, a fresh browser will trigger bot detectiomn. YOu have to browse the web first, then handle logins manually yourself, then youa re good to go for about 3 days.
This repository exists only for technical research.
This project interacts with third-party platforms in ways that may violate their Terms of Service, including:
- Automated interaction
- Scraping
- Circumventing API billing
- Browser automation that mimics human behavior
Using this against any real service can result in:
- ❌ Permanent OpenAI account bans
- ❌ Permanent LinkedIn account bans
- ❌ IP blacklisting
- ❌ Legal consequences
Do NOT use this in production.
Do NOT connect it to real accounts.
Use at your own risk.
This research prototype demonstrates:
- Browser-driven automation
- End-to-end content generation pipelines
- LLM-assisted social posting logic
- A small FastAPI service for experimentation
This is not an API client.
It does not use the official OpenAI API.
It must not be used to avoid legitimate API costs.
- 🚫 Circumventing payment systems
- 🚫 Automating real user accounts
- 🚫 Impersonating humans
- 🚫 Spamming, phishing, or mass posting
- 🚫 Growth-hacking or engagement fraud
Any misuse of this project is entirely your liability.
python \-m venv .venv
source .venv/bin/activate
pip install \-r requirements.txt uvicorn main:app \-\-reload Starts the local FastAPI server with auto-reload enabled.