Devil Agent Ai Platform

⚡ Devil Agent — Autonomous AI Engineering Operating System

🔥 The World's Most Advanced Autonomous AI Agent Platform

Built with NEXUS OS Framework | Agent Core | MCP Server Architecture

[Features](#-revolutionary-features) • [Quick Start](#-quick-start) • [Architecture](#-architecture) • [Deployment](#-deployment) • [Contributing](#-contributing)

---

🌟 Why Choose Devil Agent?

Devil Agent isn't just another AI assistant — it's a fully autonomous engineering operating system that combines cutting-edge AI capabilities with production-grade reliability. Here's what makes it truly unique:

🚀 Unmatched Capabilities

• 100% Autonomous Operation — Self-directed planning, execution, and verification
• Multi-Layer Reasoning — Strategic, Tactical, Execution, and Sentinel layers running in parallel
• Zero Regression Guarantee — Automated snapshots and rollbacks for every change
• Enterprise-Grade Security — OWASP compliance, CVE scanning, and zero-trust architecture
• Real-Time Learning — Continuous self-improvement through failure pattern analysis

💎 Rare Features No One Else Has

1. NEXUS Autonomous Engineering System

• 4-layer cognitive architecture (Strategic → Tactical → Execution → Sentinel)
• Self-healing infrastructure with automatic remediation
• Quantum-ready cryptographic assessment
• Probabilistic release readiness scoring

1. Agent Core Skill Library

• 100+ pre-built specialized skills (Finance, Research, Web Dev, Data Analysis)
• Anthropic-authored professional skills (PowerPoint, Excel, Financial Modeling)
• Dynamic skill injection and hot-swapping
• Custom skill development framework

1. MCP (Model Context Protocol) Server

• Universal AI model integration layer
• Support for Claude, GPT-4, Gemini, and custom models
• Context-aware model switching
• Built-in rate limiting and cost optimization

1. Production-Ready Full-Stack Platform

• React 19 + TypeScript frontend with shadcn/ui
• FastAPI backend with async/await patterns
• MongoDB for persistent storage
• WebSocket real-time communication
• Key-based authentication (40-word secret keys)

1. Advanced Security Features

• No username/password — only cryptographic secret keys
• Zero-knowledge architecture
• Automatic secret detection and rotation
• Supply chain security (SBOM + provenance)
• Compliance automation (SOC2, GDPR, HIPAA ready)

1. Self-Improving AI

• Failure pattern library that learns from mistakes
• Performance regression detection
• Knowledge staleness tracking
• Capability gap logging for future improvements

---

📚 Table of Contents

• [Revolutionary Features](#-revolutionary-features)
• [Quick Start](#-quick-start)
• [Architecture](#-architecture)
• [Installation](#-installation)
• [Configuration](#-configuration)
• [Usage](#-usage)
• [MCP Server](#-mcp-server)
• [Agent Core Skills](#-agent-core-skills)
• [NEXUS OS](#-nexus-os-framework)
• [Security](#-security)
• [Deployment](#-deployment)
• [API Documentation](#-api-documentation)
• [Contributing](#-contributing)
• [Troubleshooting](#-troubleshooting)
• [License](#-license)
• [Tags](#-tags)

---

🎯 Revolutionary Features

Core Platform Features

| Feature | Description | Status | |---------|-------------|--------| | Autonomous Planning | Multi-step task decomposition with dependency graphs | ✅ Production | | Self-Healing | Automatic error detection and recovery | ✅ Production | | Zero Downtime Deploys | Blue-green + canary deployment strategies | ✅ Production | | Real-Time Monitoring | Live metrics, logs, and performance tracking | ✅ Production | | Cost Optimization | Automatic rightsizing and resource optimization | ✅ Production | | Security Scanning | Continuous CVE monitoring and patching | ✅ Production | | Chaos Engineering | Automated resilience testing | 🚧 Beta | | Quantum Readiness | Post-quantum cryptography migration | 📋 Planned |

AI/ML Capabilities

• Multi-Model Support: Claude Opus 4.5, GPT-5.2, Gemini 3 Pro
• Function Calling: Advanced tool use and API integration
• RAG (Retrieval Augmented Generation): Semantic codebase search
• Code Generation: Context-aware code synthesis
• Test Generation: Automatic unit + integration test creation
• Documentation Generation: Auto-generated API docs and guides

Developer Experience

• Hot Reload: Frontend and backend auto-reload on changes
• Type Safety: Full TypeScript support with strict mode
• Linting: ESLint + Prettier with auto-fix
• Testing: Jest + React Testing Library + Playwright
• CI/CD: GitHub Actions workflows included
• Docker Support: Multi-stage builds with optimization

---

🚀 Quick Start

Prerequisites

• Node.js 18+ and npm/yarn
• Python 3.9+ and pip
• MongoDB 5.0+
• Redis 6.0+ (optional, for caching)
• Git

One-Command Installation

# Clone the repository
git clone https://github.com/vikrant-project/devil-agent-ai-platform.git
cd devil-agent-ai-platform

# Run the automated setup script
chmod +x setup.sh
./setup.sh

Manual Installation

1. Backend Setup

cd backend

# Create virtual environment
python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Copy environment variables
cp .env.example .env

# Edit .env and add your configuration
nano .env

2. Frontend Setup

cd frontend

# Install dependencies
npm install
# or
yarn install

# Copy environment variables
cp .env.example .env

# Edit .env and add your configuration
nano .env

3. Database Setup

# Start MongoDB
sudo systemctl start mongod

# Or use Docker
docker run -d -p 27017:27017 --name mongodb mongo:latest

4. Start the Application

# Terminal 1 - Backend
cd backend
source venv/bin/activate
python server.py

# Terminal 2 - Frontend
cd frontend
npm start

Visit http://localhost:3000 🎉

---

🏗️ Architecture

System Overview

┌─────────────────────────────────────────────────────────────┐
│                        Devil Agent Platform                  │
├─────────────────────────────────────────────────────────────┤
│                                                               │
│  ┌──────────────┐     ┌──────────────┐     ┌──────────────┐│
│  │   Frontend   │────▶│   Backend    │────▶│   MongoDB    ││
│  │  React + TS  │     │   FastAPI    │     │              ││
│  └──────────────┘     └──────────────┘     └──────────────┘│
│         │                     │                              │
│         │                     ▼                              │
│         │              ┌──────────────┐                      │
│         │              │  MCP Server  │                      │
│         │              └──────────────┘                      │
│         │                     │                              │
│         │                     ▼                              │
│         │              ┌──────────────┐                      │
│         └─────────────▶│  Agent Core  │                      │
│                        │  (Skills)    │                      │
│                        └──────────────┘                      │
│                                                               │
│                        ┌──────────────┐                      │
│                        │  NEXUS OS    │                      │
│                        │  Framework   │                      │
│                        └──────────────┘                      │
└─────────────────────────────────────────────────────────────┘

Technology Stack

Frontend:

• React 19 with TypeScript
• Tailwind CSS + shadcn/ui components
• React Router v7 for routing
• Axios for API communication
• Lucide React for icons
• Recharts for data visualization

Backend:

• FastAPI (Python 3.9+)
• Motor (Async MongoDB driver)
• Pydantic for validation
• JWT authentication
• bcrypt for password hashing
• emergentintegrations for AI models

Infrastructure:

• MongoDB 5.0+
• Redis (optional caching)
• Nginx (reverse proxy)
• PM2 (process management)
• Docker & Docker Compose

---

🔧 Installation

Development Environment

# Install all dependencies
npm run install:all

# Start development servers
npm run dev

# Run tests
npm run test

# Build for production
npm run build

Production Deployment

# Build optimized bundles
npm run build:prod

# Start with PM2
pm2 start ecosystem.config.js

---

⚙️ Configuration

Environment Variables

Backend (.env)

# Server Configuration
PORT=8000
HOST=0.0.0.0
ENV=production

# Database
MONGO_URL=mongodb://localhost:27017
DB_NAME=devil_agent_db

# Security
SECRET_KEY=your-super-secret-key-change-this-in-production
JWT_ALGORITHM=HS256
ACCESS_TOKEN_EXPIRE_MINUTES=60

# AI Models (Optional)
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...
GOOGLE_API_KEY=...

# Redis (Optional)
REDIS_URL=redis://localhost:6379

# CORS
CORS_ORIGINS=http://localhost:3000,https://YOUR_DOMAIN_HERE

Frontend (.env)

# API Configuration
REACT_APP_API_URL=https://YOUR_DOMAIN_HERE
REACT_APP_WS_URL=wss://YOUR_DOMAIN_HERE/ws

# Feature Flags
REACT_APP_ENABLE_MCP=true
REACT_APP_ENABLE_NEXUS=true
REACT_APP_DEBUG_MODE=false

Agent Core Configuration

Edit agent_core/DEVIL_AGENT_CONFIG.md to customize:

• Available skills
• Model preferences
• Tool permissions
• Safety constraints

---

📖 Usage

Basic Usage

from agent_core import DevilAgent

# Initialize agent
agent = DevilAgent(
    model="claude-opus-4.5",
    skills=["research", "coding", "finance"]
)

# Run autonomous task
result = agent.execute(
    task="Build a full-stack web app with user authentication",
    constraints={
        "timeline": "2 hours",
        "budget": "$0",
        "security": "high"
    }
)

print(result.summary)
print(result.artifacts)

Advanced Usage

# Multi-agent orchestration
orchestrator = AgentOrchestrator()

orchestrator.spawn_agents([
    ("architect", ArchitectAgent()),
    ("builder", BuilderAgent()),
    ("tester", TesterAgent()),
    ("security", SecurityAgent())
])

result = orchestrator.execute_workflow(
    task_graph=build_task_graph(),
    parallel_execution=True,
    max_retries=3
)

---

🔌 MCP Server

The Model Context Protocol (MCP) server provides a universal interface for AI model integration.

Starting MCP Server

cd agent_core
./start-gateway.sh

MCP Endpoints

• POST /v1/chat/completions - Chat completion
• POST /v1/embeddings - Generate embeddings
• GET /v1/models - List available models
• POST /v1/tools/execute - Execute a tool/function

Supported Models

| Provider | Models | Status | |----------|--------|--------| | Anthropic | Claude Opus 4.5, Sonnet 4.5, Haiku 4.5 | ✅ | | OpenAI | GPT-5.2, GPT-4o, GPT-4 Turbo | ✅ | | Google | Gemini 3 Pro, Gemini 3 Flash | ✅ | | Custom | Your fine-tuned models | ✅ |

---

🎓 Agent Core Skills

Agent Core includes 100+ specialized skills across multiple domains:

Finance Skills

• 3-Statement Financial Model - Build integrated P&L, Balance Sheet, Cash Flow
• DCF Valuation - Discounted Cash Flow analysis with sensitivity tables
• LBO Model - Leveraged Buyout modeling
• Comps Analysis - Comparable company analysis
• Merger Model - M&A accretion/dilution analysis
• Excel Author - Generate complex Excel spreadsheets
• PowerPoint Author - Create professional presentations

Research Skills

• Parallel CLI - Run multiple research queries in parallel
• Darwinian Evolver - Evolve solutions through genetic algorithms
• Web Scraper - Extract structured data from websites
• Literature Search - Search academic papers and publications

Web Development Skills

• Page Agent - Build full web pages from descriptions
• Component Generator - Create React/Vue/Angular components
• API Designer - Design RESTful APIs with OpenAPI specs

Productivity Skills

• Document Converter - Convert between formats (PDF, DOCX, MD)
• Data Analyzer - Analyze CSV/Excel data
• Image Processor - Resize, crop, optimize images

Adding Custom Skills

# Create a new skill
from agent_core.skills import BaseSkill

class MyCustomSkill(BaseSkill):
    name = "my-custom-skill"
    description = "Does something amazing"
    
    def execute(self, params):
        # Your logic here
        return result

# Register the skill
agent_core.register_skill(MyCustomSkill())

---

🧠 NEXUS OS Framework

NEXUS is the autonomous operating system that powers Devil Agent's decision-making.

Four Cognitive Layers

1. Strategic Mind - Long-term planning and architecture
2. Tactical Mind - Task decomposition and sequencing
3. Execution Engine - Direct action and implementation
4. Sentinel - Continuous verification and security

Core Principles

• Autonomy First - Default to autonomous execution
• Evidence-Based Completion - Tasks are "done" only with proof
• Zero Regression - Automated rollback on degradation
• Minimum Footprint - Only install what's needed
• Security by Default - OWASP Top 10 compliance

NEXUS Features

• Autonomous code refactoring
• Self-healing deployments
• Predictive technical debt tracking
• Continuous threat modeling
• Real-time cost optimization

See [NEXUSQUICKSTART.md](NEXUSQUICKSTART.md) for detailed documentation.

---

🔐 Security

Key-Based Authentication

Devil Agent uses a unique 40-word secret key system instead of traditional username/password:

Benefits:

• 🔒 Unhackable - 256-bit entropy, impossible to brute force
• 🚫 No Phishing - Keys can't be phished like passwords
• 🔑 One Key - Single key for all services
• 💾 Offline Storage - Store in password manager or encrypted file

Creating an Account:

1. Visit /signup
2. Click "Generate My Secret Key"
3. CRITICAL: Save the 40-word key immediately (it's shown only once!)
4. Store in password manager (1Password, LastPass, Bitwarden)
5. Use key to login at /login

Security Features

• Automatic Secret Scanning - Detects secrets in code/logs
• CVE Monitoring - Real-time vulnerability detection
• Zero-Trust Architecture - Every service-to-service call is authenticated
• Audit Logging - Tamper-proof logs with cryptographic signatures
• Compliance - SOC2, GDPR, HIPAA ready out of the box

Responsible Disclosure

Found a security issue? Please email: security@YOURDOMAINHERE

Do NOT open a public GitHub issue for security vulnerabilities.

---

🚀 Deployment

VPS Deployment (Ubuntu 22.04)

# 1. SSH into your VPS
ssh -i YOUR_PEM_KEY_HERE ubuntu@YOUR_VPS_IP_HERE

# 2. Clone repository
git clone https://github.com/vikrant-project/devil-agent-ai-platform.git
cd devil-agent-ai-platform

# 3. Run deployment script
chmod +x deploy.sh
./deploy.sh production

# 4. Configure Nginx
sudo cp nginx.conf /etc/nginx/sites-available/devil-agent
sudo ln -s /etc/nginx/sites-available/devil-agent /etc/nginx/sites-enabled/
sudo nginx -t && sudo systemctl reload nginx

# 5. Setup SSL with Let's Encrypt
sudo certbot --nginx -d YOUR_DOMAIN_HERE

Docker Deployment

# Build and run with Docker Compose
docker-compose up -d

# View logs
docker-compose logs -f

# Stop services
docker-compose down

Cloud Platforms

AWS:

• EC2 + RDS + S3
• Elastic Beanstalk
• ECS + Fargate

Google Cloud:

• Compute Engine + Cloud SQL
• App Engine
• Cloud Run

Azure:

• Virtual Machines + Cosmos DB
• App Service
• Container Instances

See [docs/DEPLOYMENTSUMMARY.md](docs/DEPLOYMENTSUMMARY.md) for platform-specific guides.

---

📡 API Documentation

Authenti

…

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: vikrant-project
• Source: vikrant-project/devil-agent-ai-platform
• License: MIT

Install and usage instructions live in the source repository linked above.