Cortex

🧠 AI-powered infrastructure debugging orchestrator
Organize, automate, and share debugging workflows with ease

Quick Start • Features • Architecture • Documentation • Contributing

What is Cortex?

Cortex helps teams organize debugging knowledge into reusable components called neurons (discrete tasks) and chain them into automated synapses (workflows). Think of it as bringing structure and science to the art of infrastructure debugging.

Visual Overview

flowchart LR
    A["Neurons<br/>(Debugging Tasks)"] --> B["Synapses<br/>(Workflows)"]
    B --> C["Insights<br/>(Results)"]
    D["AI Generator"] -.->|generates| A

    style A fill:#e1f5ff,stroke:#3b82f6,stroke-width:2px,color:#1f2937
    style B fill:#fff4e1,stroke:#f59e0b,stroke-width:2px,color:#1f2937
    style C fill:#c8e6c9,stroke:#10b981,stroke-width:2px,color:#1f2937
    style D fill:#f3e5f5,stroke:#8b5cf6,stroke-width:2px,color:#1f2937

Key Benefits

🧠 Knowledge Organization: Turn debugging scripts into reusable neurons
🔗 Workflow Automation: Chain neurons into powerful synapses
🤖 AI-Powered: Generate debugging scripts from natural language (coming soon)
📊 Visual Insights: Modern web UI with real-time execution tracking (coming soon)
🚀 Deploy Anywhere: Single 50MB binary OR Kubernetes cluster

Features

✅ Neuron Execution: Run discrete debugging tasks with structured output
✅ Synapse Orchestration: Chain neurons in sequential or parallel workflows
✅ CLI Interface: Powerful command-line tool with intuitive commands
🚧 AI Generation: Generate neurons from natural language (OpenAI, Anthropic, Ollama)
🚧 Web Dashboard: Real-time monitoring with WebSocket streaming
🚧 Visual Builder: Drag-and-drop synapse creation
📋 Fleet Management: Monitor distributed edge devices
📋 Plugin System: Extend functionality with custom plugins

Quick Start

Installation

# Clone the repository
git clone https://github.com/anoop2811/cortex.git
cd cortex

# Build the binary
make build

# Verify installation
./cortex --version

Your First Neuron

A neuron is a discrete debugging task (shell script + config).

# Create a neuron
cortex create-neuron check-nginx

# This creates:
# check-nginx/
# ├── config.yml
# └── run.sh

How it works:

sequenceDiagram
    participant User
    participant Cortex
    participant Neuron
    participant System

    User->>Cortex: cortex exec -p check-nginx
    Cortex->>Neuron: Load config.yml
    Neuron-->>Cortex: Configuration loaded
    Cortex->>System: Execute pre_exec_debug
    System-->>User: Checking nginx status...
    Cortex->>System: Execute run.sh
    System->>System: systemctl is-active nginx
    System-->>Cortex: Exit code 0
    Cortex->>User: Success - Nginx is running!

Edit check-nginx/config.yml:

---
name: check_nginx
type: check
description: "Check if nginx is running"
exec_file: run.sh
pre_exec_debug: "Checking nginx status..."
post_exec_success_debug: "Nginx is running!"

Edit check-nginx/run.sh:

#!/bin/bash
systemctl is-active nginx

Execute:

cortex exec -p check-nginx

Your First Synapse

A synapse is a workflow that chains multiple neurons.

# Create a synapse
cortex create-synapse health-check

Configure health-check/config.yml:

---
name: health-check
description: "Complete system health check"
neurons:
  - check-nginx
  - check-database
  - check-disk-space
execution: sequential  # or parallel
stopOnError: true

Execute:

cortex exec -p health-check

Synapse Execution Flow:

flowchart LR
    Start([Execute Synapse]) --> N1[check-nginx]
    N1 -->|Success| N2[check-database]
    N1 -->|Failure| Stop1[Stop<br/>stopOnError]
    N2 -->|Success| N3[check-disk-space]
    N2 -->|Failure| Stop2[Stop]
    N3 -->|Success| Done[All Checks Passed]
    N3 -->|Failure| Stop3[Failed]

    style Start fill:#e1f5ff,stroke:#3b82f6,stroke-width:2px,color:#1f2937
    style N1 fill:#fff4e1,stroke:#f59e0b,stroke-width:2px,color:#1f2937
    style N2 fill:#fff4e1,stroke:#f59e0b,stroke-width:2px,color:#1f2937
    style N3 fill:#fff4e1,stroke:#f59e0b,stroke-width:2px,color:#1f2937
    style Done fill:#c8e6c9,stroke:#10b981,stroke-width:2px,color:#1f2937
    style Stop1 fill:#ffcdd2,stroke:#ef4444,stroke-width:2px,color:#1f2937
    style Stop2 fill:#ffcdd2,stroke:#ef4444,stroke-width:2px,color:#1f2937
    style Stop3 fill:#ffcdd2,stroke:#ef4444,stroke-width:2px,color:#1f2937

Architecture

High-Level System Design:

graph TB
    subgraph "User Interfaces"
        CLI[CLI Interface<br/>cobra + viper]
        WebUI[Web Dashboard<br/>React + Vite]
    end

    subgraph "Core Engine"
        Orchestrator[Orchestrator]
        NeuronExec[Neuron Executor]
        SynapseDAG[Synapse DAG Engine]
        AIGen[AI Generator<br/>OpenAI/Anthropic/Ollama]
    end

    subgraph "Storage"
        FileSystem[(File System<br/>YAML + Shell)]
    end

    CLI --> Orchestrator
    WebUI -.->|future| Orchestrator
    Orchestrator --> NeuronExec
    Orchestrator --> SynapseDAG
    Orchestrator --> AIGen

    NeuronExec --> FileSystem
    SynapseDAG --> FileSystem
    AIGen -.->|future| FileSystem

    style CLI fill:#e1f5ff,color:#1f2937
    style WebUI fill:#e1f5ff,stroke-dasharray: 5 5,color:#1f2937
    style Orchestrator fill:#fff4e1,color:#1f2937
    style NeuronExec fill:#c8e6c9,color:#1f2937
    style SynapseDAG fill:#c8e6c9,color:#1f2937
    style AIGen fill:#f3e5f5,stroke-dasharray: 5 5,color:#1f2937
    style FileSystem fill:#fce4ec,color:#1f2937

📘 See detailed diagrams: Architecture Diagrams

Core Concepts

Neurons: Discrete debugging tasks (shell scripts with metadata)
Synapses: Workflows that chain neurons in DAGs
Execution: Sequential or parallel neuron execution
AI Generation: Generate neurons from natural language (coming soon)

See Architecture Documentation for details.

Documentation

📚 Complete Documentation

For Users

Getting Started Guide - Install and run your first neuron
User Guide - Complete feature documentation (coming soon)

For Contributors

Contributing Guide - How to contribute
Testing Guide - TDD workflow with Ginkgo and Playwright
Architecture Docs - System design and components

Technical Specs

AI Neuron Generation - AI feature design
Web UI Specification - Web interface design
Dependencies - All open source dependencies

Examples

Check out the examples/ directory for sample neurons and synapses:

system_health_check/ - Basic system diagnostics
k8s/k8s_cluster_health/ - Kubernetes cluster health check

Use Cases

Common Debugging Workflows:

flowchart TB
    subgraph devops["DevOps and SRE"]
        Daily[Daily Health Checks]
        Incident[Incident Response]
        Auto[Automated Remediation]
    end

    subgraph k8s["Kubernetes"]
        K8sHealth[Cluster Health]
        PodDebug[Pod Diagnostics]
        Resource[Resource Analysis]
    end

    subgraph db["Database"]
        DBHealth[Health Checks]
        Replication[Replication Status]
        Performance[Query Performance]
    end

    subgraph ai["AI-Generated"]
        PortCheck[Port Process Check]
        DiskAlert[Disk Usage Alerts]
        LogAnalysis[Log Analysis]
    end

    style Daily fill:#e1f5ff,stroke:#3b82f6,stroke-width:2px,color:#1f2937
    style Incident fill:#ffcdd2,stroke:#ef4444,stroke-width:2px,color:#1f2937
    style Auto fill:#c8e6c9,stroke:#10b981,stroke-width:2px,color:#1f2937
    style K8sHealth fill:#fff4e1,stroke:#f59e0b,stroke-width:2px,color:#1f2937
    style DBHealth fill:#f3e5f5,stroke:#8b5cf6,stroke-width:2px,color:#1f2937
    style PortCheck fill:#e8f5e9,stroke:#10b981,stroke-width:2px,color:#1f2937

DevOps & SRE

# Daily health checks
cortex exec -p daily-health-check

# Incident response
cortex exec -p incident-diagnostics

# Automated remediation
cortex exec -p auto-fix-nginx

Kubernetes Debugging

# Check cluster health
cortex exec -p k8s-cluster-health

# Debug pod issues
cortex exec -p k8s-pod-diagnostics

# Analyze resource usage
cortex exec -p k8s-resource-analysis

AI-Powered Generation (Coming Soon)

# Solve real debugging problems with natural language
cortex generate-neuron \
  --prompt "Find which process is using port 8080 and show full command with PID" \
  --provider openai

# Complex multi-step diagnostics
cortex generate-neuron \
  --prompt "Check if PostgreSQL is running, accepting connections, and responding to queries" \
  --provider anthropic

# Local generation (no API key needed)
cortex generate-neuron \
  --prompt "Show disk usage for all mounts and alert if any exceeds 80%" \
  --provider ollama

Development

Prerequisites

Go 1.25.4+
Node.js 24.x LTS (for web UI)
Make

Setup

# Install dependencies
make install-deps

# Run tests
make test-all

# Build
make build

# Run in watch mode (TDD)
make watch

Testing

We follow Test-Driven Development (TDD) with outer/inner loops:

# All tests
make test-all

# Unit tests (inner loop)
make test-unit

# Acceptance tests (outer loop)
make test-acceptance

# Coverage report
make coverage

TDD Workflow:

flowchart LR
    Red1[RED: Acceptance<br/>Test Fails] --> Red2[RED: Unit<br/>Test Fails]
    Red2 --> Green[GREEN: Write<br/>Code]
    Green --> Pass{Tests<br/>Pass?}
    Pass -->|No| Red2
    Pass -->|Yes| Refactor[REFACTOR:<br/>Improve Code]
    Refactor --> Done[Feature Complete]

    style Red1 fill:#ffcdd2,stroke:#ef4444,stroke-width:2px,color:#1f2937
    style Red2 fill:#ffcdd2,stroke:#ef4444,stroke-width:2px,color:#1f2937
    style Green fill:#c8e6c9,stroke:#10b981,stroke-width:2px,color:#1f2937
    style Refactor fill:#e1bee7,stroke:#8b5cf6,stroke-width:2px,color:#1f2937
    style Done fill:#c8e6c9,stroke:#10b981,stroke-width:2px,color:#1f2937

See Testing Guide for details.

Contributing

We welcome contributions! Please see our Contributing Guide.

Quick Start for Contributors

# 1. Fork the repository
# 2. Clone your fork
git clone https://github.com/YOUR-USERNAME/cortex.git

# 3. Create feature branch
git checkout -b feature/my-feature

# 4. Make changes and test
make test-all

# 5. Submit pull request

Ways to Contribute

🐛 Report bugs
💡 Suggest features
📝 Improve documentation
🧪 Write tests
💻 Submit pull requests
🌍 Translate documentation

Project Status

Version: 1.0 (in development)
License: Apache 2.0
Language: Go 1.25.4
Dependencies: 100% open source

Roadmap

✅ Core neuron/synapse execution
✅ CLI interface
✅ Test infrastructure (Ginkgo v2 + Playwright)
🚧 AI neuron generation (OpenAI, Anthropic, Ollama)
🚧 Web UI dashboard (React + Vite)
🚧 Visual synapse builder
🚧 Kubernetes deployment
📋 Fleet management
📋 Plugin marketplace

Technology Stack

Backend: Go 1.25.4, Cobra, Viper, Zerolog
Frontend: React 19, Vite 6, Tailwind CSS 4, TanStack Query
Testing: Ginkgo v2, Gomega, Playwright
Deployment: Docker, Kubernetes, Helm

All dependencies are verified as 100% open source. See DEPENDENCIES.md.

Why Cortex?

The Problem

Debugging infrastructure is often an art, with knowledge scattered across:

Personal scripts on laptops
Slack conversations
Tribal knowledge
"What did I do 2 weeks ago?"

The Solution

Cortex brings science to the art by:

Organizing debugging steps into reusable neurons
Automating workflows with synapses
Sharing knowledge across teams
AI-powered script generation (coming soon)

Design Principles

Edge-First: Run on anything (Raspberry Pi to Kubernetes)
Shell-Native: Embrace existing bash/PowerShell scripts
Progressive: Start simple, add features as needed
Privacy-First: AI is optional, local mode available
Open Source: 100% Apache 2.0, no bait-and-switch

Security

Cortex takes security seriously. We use automated scanning and follow secure development practices.

Security Policy: See SECURITY.md for vulnerability reporting
Automated Scanning: Dependabot, CodeQL, Trivy, Gosec
Dependency Updates: Weekly automated security patches

Found a security issue? Please report it privately - see SECURITY.md

Community

GitHub Issues: Bug reports and feature requests
GitHub Discussions: Questions and ideas
Pull Requests: Contribute code

License

Apache License 2.0 - See LICENSE for details.

Acknowledgments

Built with ❤️ by the Cortex community.

Special thanks to all contributors.

Documentation • Getting Started • Contributing • Architecture

Name		Name	Last commit message	Last commit date
Latest commit History 92 Commits
.github		.github
acceptance		acceptance
architecture		architecture
assets		assets
cmd		cmd
docs		docs
example		example
internal		internal
logger		logger
web		web
.gitignore		.gitignore
.goreleaser.yml		.goreleaser.yml
CLAUDE.md		CLAUDE.md
DOCKER.md		DOCKER.md
Dockerfile		Dockerfile
Dockerfile.ui		Dockerfile.ui
LICENSE		LICENSE
Makefile		Makefile
QUICKSTART.md		QUICKSTART.md
README.md		README.md
SECURITY.md		SECURITY.md
docker-compose.yml		docker-compose.yml
go.mod		go.mod
go.sum		go.sum
main.go		main.go

Folders and files

Latest commit

History

Repository files navigation

Cortex

What is Cortex?

Visual Overview

Key Benefits

Features

Quick Start

Installation

Your First Neuron

Your First Synapse

Architecture

Core Concepts

Documentation

For Users

For Contributors

Technical Specs

Examples

Use Cases

DevOps & SRE

Kubernetes Debugging

AI-Powered Generation (Coming Soon)

Development

Prerequisites

Setup

Testing

Contributing

Quick Start for Contributors

Ways to Contribute

Project Status

Roadmap

Technology Stack

Why Cortex?

The Problem

The Solution

Design Principles

Security

Community

License

Acknowledgments

About

Resources

License

Security policy

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages