⚠️ All Rights Reserved. This repository is published for viewing and portfolio purposes only. The code is not open source: reuse, redistribution, modification, or derivative works are not permitted without written permission. See LICENSE.

💩 BullshiftDetector

AI-powered LinkedIn cringe detector, roaster & rewriter, because someone had to do it.

🔴 Live Demo · 📖 API Docs · ⭐ Star this repo

---

What is BullshiftDetector?

BullshiftDetector is a full-stack AI application that analyzes LinkedIn posts for corporate buzzwords, hollow humility, and performative authenticity, then roasts them with Claude AI.

💡 Tired of "Humbled and excited to announce"? So are we. So we built a detector.

---

Features

• 🎯 Cringe Score (0–100) · AI-powered scoring of how much bullshift is in the post
• 💩 Buzzword Detection · highlights offending words in real time
• 🔥 The Roast · one savage AI-generated one-liner about the post
• ✅ Clean Rewrite · what they should have said like a normal human
• 📋 Copy & Share · copy the clean version or share your cringe score to LinkedIn
• 🌗 Dark / Light Mode · because we're professional like that

---

Tech Stack

Layer	Technology
🧠 AI Engine	Claude API (claude-sonnet-4-5)
⚙️ Backend	FastAPI + Python 3.12
🎨 Frontend	Next.js 14 + Tailwind CSS
☁️ Backend Deploy	GCP Cloud Run
🔥 Frontend Deploy	Firebase Hosting
📦 Container	Docker
🔁 CI/CD	GCP Cloud Build

---

How It Works

User pastes LinkedIn post
        ↓
FastAPI backend receives request
        ↓
Claude API analyzes for buzzwords, cringe score, roast & rewrite
        ↓
Returns structured JSON response
        ↓
Next.js frontend renders animated results

---

Run Locally

1. Clone

git clone https://github.com/rohithkandula19/BullshiftDetector.git
cd BullshiftDetector

2. Backend

cd backend
python3.12 -m venv venv
source venv/bin/activate
cp .env.example .env
# add your ANTHROPIC_API_KEY to .env
pip install -r requirements.txt
uvicorn app.main:app --reload --port 8080

Backend: http://localhost:8080
Swagger: http://localhost:8080/docs

3. Frontend

cd frontend
cp .env.example .env.local
# set NEXT_PUBLIC_API_URL=http://localhost:8080
npm install
npm run dev

Frontend: http://localhost:3000

---

API

POST /api/v1/detect

Request:

{
  "post": "Humbled and excited to announce..."
}

Response:

{
  "score": 89,
  "buzzwords": ["humbled", "excited to announce", "grateful"],
  "roast": "Congratulations on saying absolutely nothing in 47 words.",
  "clean": "I got a new job. Thanks everyone.",
  "verdict": "PURE UNCUT BULLSHIFT"
}

---

Deploy

Backend → GCP Cloud Run

cd backend
gcloud artifacts repositories create bullshift \
  --repository-format=docker \
  --location=us-central1

gcloud builds submit \
  --tag us-central1-docker.pkg.dev/YOUR_PROJECT_ID/bullshift/bullshift-backend

gcloud run deploy bullshift-backend \
  --image us-central1-docker.pkg.dev/YOUR_PROJECT_ID/bullshift/bullshift-backend \
  --platform managed \
  --region us-central1 \
  --allow-unauthenticated \
  --set-env-vars ANTHROPIC_API_KEY=your_key_here

Frontend → Firebase

cd frontend
npm run build
firebase deploy --only hosting

---

Project Structure

bullshiftdetector/
├── backend/
│   ├── app/
│   │   ├── main.py
│   │   ├── routes/detect.py
│   │   ├── services/claude_service.py
│   │   └── models/schemas.py
│   ├── requirements.txt
│   ├── Dockerfile
│   └── .env.example
├── frontend/
│   ├── src/
│   │   ├── app/page.tsx
│   │   ├── components/
│   │   │   ├── CringeMeter.tsx
│   │   │   └── HighlightedPost.tsx
│   │   └── lib/api.ts
│   ├── package.json
│   └── firebase.json
└── README.md

---

Why I Built This

Because I got tired of seeing "Humbled and excited to announce" for the 47th time this week.

Also, it makes recruiters laugh. And laughing recruiters give offers. 😂

---

built with 🤖 Claude API · no posts stored · roasts fictional · zero bullshift tolerance 💩