Your infrastructure never sleeps. Neither does Sentinel.

Quick Start • Features • Architecture • Hackathon • Contributing

---

🏆 Hackathon Achievement

🥇 FEATURED PROJECT at WeMakeDevs AI Agents Assemble

Metric	Value
Total Teams Competing	6,000+ worldwide
Countries Represented	20+
Prize Pool	$15,000 USD
Sentinel's Status	✨ Featured in Top Projects ✨

Submitted to: Apertre 3.0 Open Source Programme

---

🎯 What is Sentinel?

Sentinel is an autonomous AI-powered DevOps agent that transforms infrastructure management from reactive firefighting to proactive, self-healing operations. Unlike traditional monitoring tools that tell you what broke, Sentinel tells you why it broke and fixes it automatically—without human intervention. The Problem We Solve ❌ Engineers woken at 3 AM to restart services ❌ Alert fatigue from constant notifications ❌ MTTR (Mean Time To Recovery) measured in hours ❌ Post-mortem blame cycles instead of prevention Sentinel's Solution ✅ 24/7 autonomous monitoring with 5-second polling ✅ AI-powered root cause analysis (Groq LLaMA 3.3-70B) ✅ Automatic self-healing within 30 seconds ✅ Transparent reasoning for every decision ✅ Cost-optimized: AI only runs when services actually fail

┌─────────────────────────────────┐ │ 🛡️ SENTINEL ARCHITECTURE │ ├─────────────────────────────────┤ │ │ │ Your Infrastructure │ │ ┌─────┐ ┌─────┐ ┌─────┐ │ │ │ 🟢 │ │ 🟢 │ │ 🔴 │ │ │ │Auth │ │ Pay │ │Notif│ │ │ └──┬──┘ └──┬──┘ └──┬──┘ │ │ │ │ │ │ │ └───────┼────────┘ │ │ ▼ │ │ ┌──────────────┐ │ │ │ 🤖 AI Engine │ │ │ │ (Kestra + │ │ │ │ Groq) │ │ │ └──────┬───────┘ │ │ ▼ │ │ ✨ AUTO-HEAL ✨ │ │ ▼ │ │ All Services Healthy │ │ │ └─────────────────────────────────┘

---

✨ Key Features

🧠 Intelligence Layer Real-time Root Cause Analysis using LLaMA 3.3-70B Predictive Failure Detection via pattern recognition Cost-Optimized AI: Only invokes LLM when services fail Human-Readable Reports with actionable insights ⚡ Automation Layer 30-Second Kestra Orchestration for workflow automation Autonomous Self-Healing without human approval Multi-Service Monitoring with parallel health checks Intelligent Recovery Workflows tailored to failure types

🎛️ Visibility Layer Real-Time Dashboard with live metrics & service status AI Reasoning Panel showing agent's decision-making Incident Timeline with recovery analytics CLI Tool for power users and developers 🐳 Infrastructure Layer Docker Containerization for easy deployment PostgreSQL State Management via Kestra Webhook-Based Communication between components Scalable Microservices Architecture for production use

---

🖼️ System in Action

📊 Real-Time Dashboard

All systems healthy with live metrics and AI reasoning

🚨 Detecting Failures

Service down detected, AI analysis triggered

🤖 CLI Power Tool

Developer interface for manual operations

---

🏗️ Architecture

┌──────────────────────────────────────────────────────────────────────────┐
│                        🛡️ SENTINEL STACK                                 │
├──────────────────────────────────────────────────────────────────────────┤
│                                                                          │
│  ┌────────────────┐    ┌────────────────┐    ┌────────────────┐        │
│  │   📱 FRONTEND  │    │  🔧 BACKEND    │    │  🤖 KESTRA     │        │
│  │  Next.js 16    │◄──►│  Express.js    │◄──►│  Orchestrator  │        │
│  │  Port: 3000    │    │  Port: 4000    │    │  Port: 9090    │        │
│  └────────────────┘    └────────┬───────┘    └────────┬───────┘        │
│                                 │                     │                 │
│                    ┌────────────┴─────────────────────┘                 │
│                    ▼                                                    │
│   ┌─────────────────────────────────────────────────────────────────┐  │
│   │         🐳 DOCKER NETWORK (Services + State)                    │  │
│   ├─────────────────────────────────────────────────────────────────┤  │
│   │                                                                 │  │
│   │  ┌──────────┐   ┌──────────┐   ┌──────────┐   ┌─────────────┐ │  │
│   │  │ 🔐 Auth  │   │ 💳 Pay   │   │ 📧 Notif │   │ 🗄️ Postgres  │ │  │
│   │  │  :3001   │   │  :3002   │   │  :3003   │   │   :5432     │ │  │
│   │  └──────────┘   └──────────┘   └──────────┘   └─────────────┘ │  │
│   │                                                                 │  │
│   └─────────────────────────────────────────────────────────────────┘  │
│                                                                          │
│                            ┌──────────┐                                 │
│                            │ 🧠 GROQ  │                                 │
│                            │ LLaMA AI │                                 │
│                            └──────────┘                                 │
│                                                                          │
└──────────────────────────────────────────────────────────────────────────┘

How It Works

sequenceDiagram
    participant B as Backend<br/>(5s polling)
    participant K as Kestra<br/>(30s cron)
    participant S as Services
    participant AI as Groq<br/>LLaMA 3.3
    participant F as Dashboard

    loop Every 5 Seconds
        B->>S: Health check all services
        S-->>B: Status responses
        B->>F: Broadcast status
    end

    loop Every 30 Seconds
        K->>S: Parallel health checks
        S-->>K: Responses
        
        alt Any Service Down?
            K->>AI: Analyze failure + metrics
            AI-->>K: Root cause + recommendations
            K->>S: Execute healing action
            K->>B: POST webhook with report
            B->>F: Real-time update
        else All Healthy
            K->>B: Send healthy status
        end
    end

Loading

Recovery Timeline

Time	Event
0s	Service crashes
5s	Backend detects (5-second polling)
30s	Kestra scheduled check runs
32s	AI analyzes root cause
35s	Healing action executes
40s	Service restored ✅

Worst-case recovery: ~65 seconds (when service fails just after Kestra check)

---

🛠️ Tech Stack

Layer	Tech	Purpose
Frontend	Next.js 16, TypeScript, Tailwind CSS, Recharts	Real-time dashboard with glassmorphism UI
Backend	Node.js, Express, Axios	Health aggregation, webhook handler, REST API
Orchestration	Kestra, YAML workflows, PostgreSQL	Automation, state management, scheduling
AI/Intelligence	Groq API (LLaMA 3.3-70B)	Root cause analysis, recommendations
Infrastructure	Docker, Docker Compose, 3 Mock Services	Containerization, networking, simulation
CLI	Commander.js, Chalk, cli-table3	Developer interface, chaos testing

---

🚀 Quick Start

Prerequisites

• Docker & Docker Compose (v20+)
• Node.js 18+ (for local development)
• PostgreSQL 12+ (required for RBAC system - included in docker-compose)
• Groq API Key (Free at console.groq.com)

Note: The RBAC system requires PostgreSQL. When using docker-compose up, PostgreSQL is automatically started. For local development, install PostgreSQL separately or use the containerized version.

⚡ One-Command Setup

# Clone the repository
git clone https://github.com/SKfaizan-786/sentinel-devops-agent.git
cd sentinel-devops-agent

# Set up environment variables
cp backend/.env.example backend/.env
# ⚠️  Edit backend/.env and set a strong JWT_SECRET before starting!

# Start the entire stack (includes PostgreSQL for RBAC)
docker-compose up -d

# Initialize RBAC system (first time only)
cd backend
npm install
npm run quick-setup

# That's it! Access at:
# 🌐 Dashboard: http://localhost:3000
# 🤖 Kestra UI: http://localhost:9090
# 📊 Backend API: http://localhost:4000

⚠️ Security Warning: The quick-setup creates a default admin account (admin@example.com / password123) for development. Change this password immediately in production environments!

🔧 Development Setup

Expand for full development guide

# 1. Start infrastructure (includes PostgreSQL for RBAC)
docker-compose up -d kestra postgres auth-service payment-service notification-service

# 2. Set up RBAC system
cd backend
cp .env.example .env
# ⚠️  Edit .env and set JWT_SECRET to a strong random value:
# node -e "console.log(require('crypto').randomBytes(32).toString('hex'))"
npm install
npm run quick-setup  # Creates database schema and default admin

# 3. Start backend (in same terminal)
npm start

# 4. Start frontend (in new terminal)
cd ../sentinel-frontend
npm install
npm run dev

# 5. Optional: Install CLI
cd ../cli
npm install
npm link

# Now accessible at:
# Dashboard: http://localhost:3000
# Backend: http://localhost:4000
# Kestra: http://localhost:9090
# PostgreSQL: localhost:5432 (for RBAC)
# CLI: sentinel status

# ⚠️  Default admin credentials (DEVELOPMENT ONLY):
# Email: admin@example.com
# Password: password123
# Change immediately in production!

---

🖥️ CLI Usage

Sentinel includes a powerful CLI for DevOps engineers:

# View system health
sentinel status

# Simulate failures (chaos testing)
sentinel simulate auth down
sentinel simulate payment degraded
sentinel simulate notification slow

# Trigger manual healing
sentinel heal auth

# Generate AI incident report
sentinel report

Example Output:

$ sentinel status

🛡️  SENTINEL STATUS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

  Service              Status    Latency   
  ─────────────────────────────────────────
  auth-service         🟢 UP     45ms      
  payment-service      🟢 UP     52ms      
  notification-service 🟢 UP     38ms      

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Last Check: 2026-02-01T10:30:00Z
AI Status: Active & Monitoring

---

🧪 Test the Auto-Healing

Live Demo Scenario:

# Terminal 1: Watch the dashboard
open http://localhost:3000

# Terminal 2: Crash a service
sentinel simulate auth down

# Watch what happens:
# 1. Dashboard status → RED (within 5s)
# 2. AI panel → "Analyzing..."
# 3. After 30s → Kestra runs + triggers healing
# 4. Within 65s total → Service restored, status → GREEN

---

🏆 Hackathon Qualifications

Track	Achievement
🤖 Kestra	Autonomous Kestra orchestration with parallel health checks, conditional AI invocation, and self-healing workflows
💻 Cline	Built with Cline's assistance. Production-ready CLI embodies autonomous developer workflows
🐰 CodeRabbit	AI-powered code reviews on every PR ensure enterprise-grade quality
🌐 Vercel	Real-time dashboard deployed on Vercel with optimized Next.js

---

📚 Documentation

Document	Content
DOCUMENTATION.md	Complete docs index
ARCHITECTURE.md	System design deep-dive
DEVELOPMENT.md	Setup & development guide
CONTRIBUTING.md	How to contribute
SECURITY.md	Security policy & disclosure
API.md	REST API reference
FAQ.md	50+ Q&A
ROADMAP.md	Future features

---

🤝 Contributing

We welcome contributions! Sentinel is open source and beginner-friendly.

# Fork → Clone → Branch → Code → Push → PR
git checkout -b feature/amazing-feature
git commit -m "feat: add amazing feature"
git push origin feature/amazing-feature

See CONTRIBUTING.md and CODE_OF_CONDUCT.md.

---

📄 License

MIT License - see LICENSE for details.

---

👥 Team

@SKfaizan-786 Backend & Orchestration

@mdhaarishussain Frontend & Dashboard

Built with ❤️ for the WeMakeDevs AI Agents Assemble Hackathon
Featured in the Top Projects (6000+ teams worldwide)

---

⭐ Show Your Support

If Sentinel helped you, give us a star! ⭐

Share with your network:

---

Sentinel
"Monitoring that never sleeps"

Made with 🛡️ by the Sentinel Team