devops-sentinel 0.1.1

Autonomous SRE agents that watch your services, detect anomalies, and generate blameless postmortems

DevOps Sentinel

Autonomous SRE Agent System - Reduce Mean Time to Detection (MTTD) from minutes to under 10 seconds.

Features

• Continuous Monitoring - Health checks at configurable intervals
• Multi-Agent Analysis - CrewAI-powered investigation and root cause analysis
• AI Postmortems - Automated incident documentation
• Real-time Dashboard - WebSocket-powered live updates
• CLI Interface - Similar to Gemini CLI
• Pre-Deploy Linting - Code quality checks before production
• Privacy-First - No telemetry, transparent data handling

Architecture

┌─────────────────────────────────────────────────────────────┐
│                      DevOps Sentinel                         │
├─────────────────────────────────────────────────────────────┤
│  CLI (Typer)  │  API (FastAPI)  │  Dashboard (WebSocket)    │
├─────────────────────────────────────────────────────────────┤
│                   Monitoring Orchestrator                    │
├─────────────────────────────────────────────────────────────┤
│  Watcher Agent  │  Triage Agent  │  Investigator  │  Strategist  │
├─────────────────────────────────────────────────────────────┤
│  Health Check  │  Slack Alert  │  Log Analysis  │  DB Health  │
├─────────────────────────────────────────────────────────────┤
│  Supabase (Persistence)  │  OpenRouter (LLM)  │  Slack       │
└─────────────────────────────────────────────────────────────┘

Installation

# Install from PyPI (recommended)
pip install devops-sentinel

For Developers

# Clone the repository
git clone https://github.com/frozen8569/devops-sentinel.git
cd devops-sentinel

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

# Install in editable mode for development
pip install -e .

Quick Start

# Login to DevOps Sentinel
sentinel login

# Check a service health
sentinel health https://api.example.com

# Monitor a service continuously
sentinel monitor https://api.example.com/health

# View your projects
sentinel projects list

CLI Commands

Command	Description
sentinel init	Interactive configuration setup
sentinel monitor <url>	Monitor a service URL
sentinel status	Show monitored services
sentinel incidents	List recent incidents
sentinel postmortem <id>	View incident postmortem
sentinel serve	Start dashboard server
sentinel lint <path>	Run pre-deploy checks

Configuration

Create a .env file or run sentinel init:

OPENROUTER_API_KEY=your_key_here
DEFAULT_MODEL=google/gemini-pro
SUPABASE_URL=https://your-project.supabase.co
SUPABASE_KEY=your_anon_key
SLACK_WEBHOOK_URL=https://hooks.slack.com/...

Docker Deployment

# Build and run
docker-compose up -d

# View logs
docker-compose logs -f sentinel

How It Works

1. Watcher Agent continuously monitors service health endpoints
2. On failure, First Responder sends a Slack alert
3. Investigator analyzes logs, deployments, and database status
4. Strategist creates a prioritized action plan
5. AI generates a structured postmortem saved to Supabase

Privacy

• All data stays local or in YOUR Supabase instance
• No telemetry or external data collection
• Use --privacy flag to see exactly what data is processed
• OpenRouter calls only contain health check context (no PII)

Supabase Schema

-- Run these in your Supabase SQL editor
CREATE TABLE services (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    name TEXT NOT NULL,
    url TEXT NOT NULL,
    check_interval INTEGER DEFAULT 10,
    is_active BOOLEAN DEFAULT true,
    created_at TIMESTAMPTZ DEFAULT now()
);

CREATE TABLE incidents (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    service_id UUID REFERENCES services(id),
    status TEXT DEFAULT 'detecting',
    detected_at TIMESTAMPTZ DEFAULT now(),
    resolved_at TIMESTAMPTZ,
    mttd_seconds FLOAT,
    postmortem TEXT
);

CREATE TABLE health_checks (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    service_id UUID REFERENCES services(id),
    status_code INTEGER,
    response_time_ms FLOAT,
    is_healthy BOOLEAN,
    checked_at TIMESTAMPTZ DEFAULT now()
);

License

MIT License