Skip to content

rohanjain11/AgentSquared

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 

Repository files navigation

<<<<<<< HEAD

Agent Squared

A no-code platform for creating config-driven AI business agents, powered by Gemini.

Quick Start

Backend (FastAPI)

cd backend
python -m venv venv
venv\Scripts\activate       # Windows
pip install -r requirements.txt
cp .env.example .env        # Add your GEMINI_API_KEY
uvicorn main:app --reload

Backend runs at http://localhost:8000. API docs at http://localhost:8000/docs.

Frontend (Next.js)

cd frontend
npm install                 # Already done by create-next-app
npm run dev

Frontend runs at http://localhost:3000.

Agent Types

Type Description Knowledge
support_qa Customer support Q&A agent RAG over uploaded docs
social_marketing Social media content generator Business context in prompt

Architecture

One runtime, two templates.
Every agent = DB row + JSON config + optional files.
agent_type selects prompt template + knowledge strategy.
No code is generated per agent.

Project Structure

Agent-squared/
├── backend/          FastAPI + SQLite + Gemini services
│   ├── routers/      API endpoints
│   ├── services/     Gemini spec gen, chat, extraction, RAG
│   ├── templates/    Agent type configurations
│   └── db/           SQLAlchemy models
└── frontend/         Next.js App Router
    └── src/app/      Pages: /, /build, /success, /a/[slug]
=======
# Agent Squared 🦋

A no-code platform for creating config-driven AI business agents, powered by Google Gemini. Monitor your brand on Bluesky, answer customer questions, and automate your business workflows without writing a single line of code.

---

## 🚀 Quick Start

### 1. Backend (FastAPI)

The backend handles the AI logic, Bluesky interactions, and RAG (Retrieval Augmented Generation).

```bash
cd backend
python3 -m venv venv
source venv/bin/activate       # macOS/Linux
# venv\Scripts\activate        # Windows

pip install -r requirements.txt

# Initial database setup
export PYTHONPATH=$PYTHONPATH:.
python3 db/init_db.py

# Run the server
uvicorn main:app --reload --port 8000

Environment Variables (.env): Create a .env file in the backend/ directory:

GEMINI_API_KEY=your_gemini_key
BLUESKY_HANDLE=yourname.bsky.social
BLUESKY_PASSWORD=your-app-password

2. Frontend (Next.js)

The frontend provides a sleek, modern dashboard for managing your AI agents.

cd frontend
npm install
npm run dev

Visit http://localhost:3000 to start building.


🛠 Agent Types

Type Description Knowledge Strategy
Support QA Automated customer help desk RAG (Upload PDFs/Txt)
Social Monitor Real-time Bluesky brand management Direct API + Sentiment Analysis

🦋 Bluesky Integration

Agent Squared is fully integrated with the AT Protocol.

  1. Real-time Monitoring: Scans your notifications for mentions and replies.
  2. Sentiment Analysis: Uses Gemini to classify mentions as Praise, Question, Complaint, or Spam.
  3. One-Click Approval: Approve AI-generated suggested replies to post them live to Bluesky.
  4. Threaded Replies: Automatically joins existing conversations with proper threading.

📁 Project Structure

AgentSquared/
├── backend/          FastAPI + SQLite + Gemini
│   ├── routers/      Chat, Social, and Agent management
│   ├── services/     Bluesky Client, Gemini Chat, RAG
│   ├── templates/    Agent behavior configurations
│   └── db/           SQLAlchemy models & migrations
└── frontend/         Next.js 16 + Tailwind CSS
    └── src/app/      Dashboard, Build, and Social views
>>>>>>> 6f1ce0f (Checkpoint)

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors