Skip to content

Alenryuichi/atoms-plus

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

200 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

Atoms Plus

Atoms Plus

πŸš€ AI-Powered Low-Code Development Platform

Features β€’ Quick Start β€’ Race Mode β€’ Agent Roles β€’ Orchestrator β€’ Deployment


✨ What is Atoms Plus?

Atoms Plus is a next-generation AI development platform that combines multi-model racing, specialized agent roles, and intelligent task orchestration to supercharge your development workflow.

πŸ’‘ Think of it as your AI development team - with a Product Manager, Architect, Engineers, and more, all working together on your tasks.


🎯 Features

🏎️ Race Mode - Multi-Model Competition

Race multiple AI models simultaneously and pick the best result:

Provider Models
Alibaba Cloud Qwen Plus, Qwen Max, Qwen Turbo
DeepSeek DeepSeek V3, DeepSeek Coder
Zhipu AI GLM-4 Plus, GLM-4 Flash, GLM-4
Anthropic Claude Sonnet 4, Claude Opus 4, Claude 3.5 Sonnet
OpenAI GPT-4o, GPT-4o Mini, GPT-4 Turbo
Google Gemini 2.0 Flash, Gemini 1.5 Pro
Mistral Mistral Large

πŸ‘₯ Agent Roles - Your AI Team

8 specialized AI agents, each with unique capabilities:

Role Name Specialty
🎯 Team Leader Mike Coordination, delegation, review
πŸ“‹ Product Manager Emma Requirements, user stories, PRD
πŸ—οΈ Architect Alex System design, API design, tech selection
πŸ“Š Project Manager Sarah Task breakdown, timeline, risk assessment
πŸ’» Engineer Bob Coding, debugging, testing
πŸ“ˆ Data Analyst Diana Data analysis, visualization, ML
πŸ”¬ Deep Researcher Ryan Research, synthesis, documentation
πŸ” SEO Specialist Sophie Keyword research, content optimization

🎯 Orchestrator - Multi-Agent Coordination

Dispatch complex tasks to multiple agents working in parallel:

Your Task β†’ Auto-Analysis β†’ Multi-Role Dispatch β†’ Parallel Execution β†’ Aggregated Result
                    ↓
        β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”
        ↓           ↓           ↓
   Product Mgr   Architect   Engineer
        ↓           ↓           ↓
   Requirements  Design     Implementation
        β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
                    ↓
            Complete Solution

πŸ’³ Credits System

  • Real-time cost tracking per model
  • Token usage statistics
  • Balance management via Supabase

πŸ› οΈ Developer Tools

  • Terminal - Execute commands directly
  • Browser - Automated web interactions
  • Code Editor - Full editing with diff view
  • Git Integration - Branch, commit, push, PR
  • File Management - Browse and edit project files

πŸš€ Quick Start

Prerequisites

  • Node.js 22+
  • Python 3.12+
  • Poetry

Installation

# Clone the repository
git clone https://github.com/Alenryuichi/atoms-plus.git
cd atoms-plus

# Install dependencies
make build

# Start the application
make run

Environment Variables

# Backend (.env)
LLM_API_KEY=your-api-key
LLM_BASE_URL=https://dashscope.aliyuncs.com/compatible-mode/v1
LLM_MODEL=qwen-plus

# Frontend (.env)
VITE_BACKEND_BASE_URL=your-backend-url
VITE_SUPABASE_URL=your-supabase-url
VITE_SUPABASE_ANON_KEY=your-supabase-key

πŸ“‘ API Reference

Race Mode API

# Start a race
POST /api/v1/race/start
{
  "prompt": "Write a React component",
  "models": ["gpt-4o", "claude-sonnet-4", "qwen-plus"]
}

# Get available models
GET /api/v1/race/models

Orchestrator API

# Dispatch to multiple roles
POST /api/v1/orchestrator/dispatch
{
  "subtasks": [
    {"role": "product_manager", "task": "Write requirements"},
    {"role": "architect", "task": "Design system"},
    {"role": "engineer", "task": "Implement feature"}
  ],
  "parallel": true
}

# Get role suggestion
POST /api/v1/orchestrator/suggest-role
{
  "task": "Build a REST API for user management"
}

Roles API

# List all roles
GET /api/v1/roles/

# Get role details
GET /api/v1/roles/{role_id}

# Get role system prompt
GET /api/v1/roles/{role_id}/prompt

🌐 Deployment

Frontend (Vercel)

cd frontend
npm run build
npx vercel deploy --prod

Backend (Railway)

# Set start command
uvicorn atoms_plus.atoms_server:app --host 0.0.0.0 --port $PORT

πŸ“Š Architecture

atoms-plus/
β”œβ”€β”€ atoms_plus/           # Core extension
β”‚   β”œβ”€β”€ race_mode/        # Multi-model racing
β”‚   β”œβ”€β”€ roles/            # Agent role system
β”‚   β”œβ”€β”€ orchestrator/     # Multi-agent coordination
β”‚   └── atoms_server.py   # Server entry
β”œβ”€β”€ frontend/             # React frontend
β”‚   β”œβ”€β”€ src/components/   # UI components
β”‚   └── src/api/          # API services
└── ...


πŸ—ΊοΈ Roadmap

  • Race Mode - Multi-model competition
  • Agent Roles - 8 specialized AI agents
  • Orchestrator - Multi-agent coordination
  • Credits System - Supabase integration
  • One-Click Deployment - Auto-hosting
  • Stripe Integration - Payment processing
  • Team Collaboration - Multi-user workspaces

πŸ”— Links


🀝 Contributing

We welcome contributions! Please see our contributing guidelines.


πŸ“„ License

MIT License - see LICENSE for details.


Made with ❀️ by the Atoms Plus Team

About

AI-Powered Low-Code Development Platform

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors