GoBE - Modular & Secure Back-end

Code Fast. Own Everything. — A modular, secure, and zero-config backend for modern Go applications with full AI integration.

TL;DR

GoBE is a modular Go backend that runs with zero configuration and provides ready-to-use REST APIs, AI providers, webhooks, and MCP (Model Context Protocol). One command = a complete server with authentication, database, 4 AI providers, and integrated system tools.

make build && ./gobe start  # Zero config, instant backend with AI
curl http://localhost:3666/mcp/tools  # MCP tools ready
curl http://localhost:3666/providers  # OpenAI, Anthropic, Gemini, Groq ready
curl http://localhost:3666/v1/webhooks/health  # Webhook system ready

About the Project

GoBE is a modular backend built with Go that embodies the Kubex principle: No Lock-in. No Excuses. It delivers security, automation, and flexibility in a single binary that runs anywhere — from your laptop to enterprise clusters.

Mission Alignment

Following Kubex's mission to democratize modular technology, GoBE provides:

DX First: One command starts everything — server, database, authentication, MCP tools
Total Accessibility: Runs without Kubernetes, Docker, or complex setup
Module Independence: Every component (CLI/HTTP/Jobs/Events) is a full citizen

Current Status - Production Ready

✅ Zero-config: Auto-generates certificates, passwords, keyring storage ✅ AI Providers: OpenAI, Anthropic Claude, Google Gemini, Groq with streaming support ✅ MCP Protocol: Model Context Protocol with dynamic tool registry and shell commands ✅ Webhook System: Functional webhook processing with AMQP integration and retry logic ✅ Discord Integration: Real MCP tool execution via Discord bot with formatted responses ✅ Modular Architecture: Clean interfaces, exportable via factory/ ✅ Database Integration: PostgreSQL/SQLite via gdbase Docker management ✅ REST API: Authentication, users, products, clients, jobs, webhooks, AI chat ✅ Security Stack: Dynamic certificates, JWT, keyring, rate limiting, whitelisted shell commands ✅ CLI Interface: Complete management via Cobra commands ✅ Multi-platform: Linux, macOS, Windows (AMD64, ARM64) ✅ Testing: Unit tests + integration tests for MCP endpoints and providers ✅ CI/CD: Automated builds and releases

Project Evolution

GoBE has evolved from a simple backend server to a comprehensive AI-integrated platform. Version 1.3.5 represents a significant milestone with the addition of:

Recent Major Additions (v1.3.5)

🤖 Full AI Provider Ecosystem: Complete integration with 4 major AI providers
🔗 Real Discord MCP Integration: Transform Discord bots from placeholder to functional AI tools
📬 Production Webhooks: Fully functional webhook system with persistence and retry logic
⚡ Streaming AI Responses: Real-time AI chat with Server-Sent Events
🔧 Enhanced MCP Tools: Shell command execution with security whitelisting

The system now serves as a complete AI backend solution, enabling developers to build AI-powered applications without complex infrastructure setup. The modular architecture makes it suitable for everything from personal projects to enterprise applications.

Features

🤖 AI Integration

✨ 4 AI Providers Ready

OpenAI (GPT-3.5, GPT-4 series) with customizable base URLs
Anthropic Claude (3.5 Sonnet, Opus, Haiku) with streaming support
Google Gemini (1.5 Pro, Flash) with proper cost calculation
Groq (Llama, Mixtral models) with ultra-fast inference

🎯 Smart Provider Management

Dynamic provider switching and availability checks
External API key support per request
Automatic cost estimation and usage tracking
Streaming responses with Server-Sent Events

🔧 System Integration

✨ MCP Protocol (Model Context Protocol)

Dynamic tool registry with thread-safe execution
Built-in system monitoring tools
Shell command execution with security whitelisting
Discord bot integration with real tool functionality

📬 Production Webhook System

AMQP/RabbitMQ integration for async processing
Persistent webhook storage with retry logic
Specialized handlers (GitHub, Discord, Stripe, etc.)
RESTful management API with pagination

🏗️ Core Platform

✨ Fully modular

All logic follows well-defined interfaces, ensuring encapsulation
Can be used as a server or as a library/module
Factory pattern for all major components

🔒 Zero-config, but customizable

Runs without initial configuration, but supports customization via files
Automatically generates certificates, passwords, and secure settings
Environment variable override support

🔗 Direct integration with gdbase

Database management via Docker
Automatic optimizations for persistence and performance
Multi-database support (PostgreSQL, SQLite)

🛡️ Advanced authentication

Dynamically generated certificates
Random passwords and secure keyring
JWT token management with refresh logic

🌐 Comprehensive REST API

AI chat endpoints with streaming
Webhook management and monitoring
Authentication, user management, products, clients
System health and metrics

📋 Enterprise-grade monitoring

Protected routes, secure storage, and request monitoring
Real-time system metrics via MCP tools
Connection health checks (DB, AMQP, webhooks)

🧑‍💻 Powerful CLI

Commands to start, configure, and monitor the server
Zero-config startup with detailed logging

How to Run

One Command. All the Power.

Quick Start

# Clone and build
git clone https://github.com/kubex-ecosystem/gobe.git
cd gobe
make build

# Start everything (zero config)
./gobe start

# Server ready at http://localhost:3666
# MCP endpoints: /mcp/tools, /mcp/exec
# Health check: /health

Requirements

Go 1.25+ (for building from source)
Docker (optional, for advanced database features)
API Keys (optional, for AI providers - can be set per request)
No Kubernetes, no complex setup required

Build Options

make build          # Production build
make build-dev      # Development build
make install        # Install binary + environment setup
make clean          # Clean artifacts
make test           # Run all tests

AI Providers

GoBE includes 4 production-ready AI providers with streaming support and cost tracking.

Available Providers

Provider	Models	Features	Pricing
OpenAI	GPT-3.5, GPT-4, GPT-4o	Streaming, custom base URL	$0.002-$0.03 per 1K tokens
Anthropic	Claude 3.5 Sonnet, Opus, Haiku	Streaming, long context	$0.25-$75 per 1M tokens
Google Gemini	Gemini 1.5 Pro, Flash	Streaming, multimodal	$0.075-$10.50 per 1M tokens
Groq	Llama 3.1, Mixtral	Ultra-fast inference	$0.05-$0.79 per 1M tokens

Usage Examples

List Available Providers

curl http://localhost:3666/providers

Chat with Streaming

curl -X POST http://localhost:3666/chat \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "openai",
    "model": "gpt-4",
    "messages": [{"role": "user", "content": "Hello!"}],
    "stream": true,
    "temperature": 0.7
  }'

Provider Switching

# Use different providers for different tasks
curl -X POST http://localhost:3666/chat \
  -d '{"provider": "anthropic", "model": "claude-3-5-sonnet-20241022", ...}'

curl -X POST http://localhost:3666/chat \
  -d '{"provider": "groq", "model": "llama-3.1-70b-versatile", ...}'

External API Keys

# Use your own API key for a specific request
curl -X POST http://localhost:3666/chat \
  -H "X-External-API-Key: your-api-key-here" \
  -d '{"provider": "openai", ...}'

Provider Configuration

Configure providers via environment variables or config files:

# Environment Variables
export OPENAI_API_KEY="sk-..."
export ANTHROPIC_API_KEY="sk-ant-..."
export GEMINI_API_KEY="..."
export GROQ_API_KEY="gsk_..."

# Custom Base URLs (for OpenAI-compatible endpoints)
export OPENAI_BASE_URL="https://api.groq.com"

Cost Tracking

Each response includes cost estimation:

{
  "content": "Response text...",
  "usage": {
    "prompt_tokens": 10,
    "completion_tokens": 20,
    "total_tokens": 30,
    "latency_ms": 1500,
    "cost_usd": 0.0006,
    "provider": "openai",
    "model": "gpt-4"
  }
}

MCP Support

GoBE implements the Model Context Protocol (MCP) for seamless AI tool integration.

Available Endpoints

GET  /mcp/tools     # List available tools
POST /mcp/exec      # Execute tools

Built-in Tools

Tool	Description	Args	Features
`system.status`	Comprehensive system status with metrics	`detailed: boolean`	Runtime stats, memory usage, connection health
`shell.command`	Execute safe shell commands	`command: string, args: array`	Whitelisted commands, 10s timeout, output capture

Example Usage

List Available Tools

curl http://localhost:3666/mcp/tools

System Status (Basic)

curl -X POST http://localhost:3666/mcp/exec \
  -H "Content-Type: application/json" \
  -d '{"tool": "system.status", "args": {"detailed": false}}'

System Status (Detailed)

curl -X POST http://localhost:3666/mcp/exec \
  -H "Content-Type: application/json" \
  -d '{"tool": "system.status", "args": {"detailed": true}}'

Execute Shell Commands

# List files
curl -X POST http://localhost:3666/mcp/exec \
  -H "Content-Type: application/json" \
  -d '{"tool": "shell.command", "args": {"command": "ls", "args": ["-la"]}}'

# Check system info
curl -X POST http://localhost:3666/mcp/exec \
  -H "Content-Type: application/json" \
  -d '{"tool": "shell.command", "args": {"command": "uname", "args": ["-a"]}}'

# Check disk usage
curl -X POST http://localhost:3666/mcp/exec \
  -H "Content-Type: application/json" \
  -d '{"tool": "shell.command", "args": {"command": "df", "args": ["-h"]}}'

Security Features

Whitelisted Commands: Only safe commands are allowed (ls, pwd, date, uname, etc.)
Timeout Protection: All commands timeout after 10 seconds
Error Handling: Proper error capture and reporting
Admin Authentication: Shell commands require admin privileges

MCP Architecture

Dynamic Registry: Tools registered at runtime, no restart needed
Thread-Safe: Concurrent tool execution with RWMutex protection
Extensible: Add custom tools via the Registry interface
Integrated: Uses existing manage controllers for system data

Webhooks

GoBE provides a production-ready webhook system with persistence, retry logic, and AMQP integration.

Features

✅ Persistent Storage: In-memory storage with plans for database persistence
✅ AMQP Integration: Async processing via RabbitMQ
✅ Retry Logic: Automatic retry of failed webhook events
✅ Specialized Handlers: GitHub, Discord, Stripe, generic webhooks
✅ RESTful API: Complete CRUD operations with pagination
✅ Real-time Stats: Monitor webhook processing in real-time

Available Endpoints

Method	Endpoint	Description
`POST`	`/v1/webhooks`	Receive webhook events
`GET`	`/v1/webhooks/health`	Webhook system health + stats
`GET`	`/v1/webhooks/events`	List webhook events (paginated)
`GET`	`/v1/webhooks/events/:id`	Get specific webhook event
`POST`	`/v1/webhooks/retry`	Retry all failed webhook events

Usage Examples

Receive Webhooks

# Generic webhook
curl -X POST http://localhost:3666/v1/webhooks \
  -H "Content-Type: application/json" \
  -H "X-Webhook-Source: github" \
  -H "X-Event-Type: push" \
  -d '{"repository": "my-repo", "commits": [...]}'

Check System Health

curl http://localhost:3666/v1/webhooks/health

Response:

{
  "status": "ok",
  "timestamp": "2025-01-20T12:00:00Z",
  "stats": {
    "total_events": 150,
    "processed_events": 145,
    "failed_events": 3,
    "pending_events": 2,
    "uptime_seconds": 3600,
    "amqp_connected": true
  }
}

List Webhook Events

# All events (paginated)
curl "http://localhost:3666/v1/webhooks/events?limit=10&offset=0"

# Filter by source
curl "http://localhost:3666/v1/webhooks/events?source=github&limit=20"

Get Event Details

curl http://localhost:3666/v1/webhooks/events/123e4567-e89b-12d3-a456-426614174000

Retry Failed Events

curl -X POST http://localhost:3666/v1/webhooks/retry

Response:

{
  "status": "success",
  "retried_count": 3,
  "timestamp": "2025-01-20T12:00:00Z",
  "message": "failed events queued for retry"
}

Webhook Event Types

The system provides specialized handling for different webhook types:

Source	Event Type	Handler Features
GitHub	`github.push`	Repository info, commit details, AMQP notifications
Discord	`discord.message`	Channel logging, bot response triggers
Stripe	`stripe.payment`	Payment processing, user billing updates
Generic	`user.created`	Welcome emails, profile creation

Integration Examples

GitHub Integration

# Configure GitHub webhook to point to your GoBE instance
# Webhook URL: https://your-server.com/v1/webhooks
# Content-Type: application/json
# Events: push, pull_request, issues

curl -X POST https://your-server.com/v1/webhooks \
  -H "X-GitHub-Event: push" \
  -H "X-Webhook-Source: github" \
  -d @github_push_payload.json

Discord Bot Integration

# Discord events are automatically processed and can trigger MCP tools
curl -X POST http://localhost:3666/v1/webhooks \
  -H "X-Webhook-Source: discord" \
  -H "X-Event-Type: message" \
  -d '{"channel_id": "123", "content": "!system status", "author": {...}}'

AMQP Integration

Webhooks are automatically published to RabbitMQ queues:

Exchange: gobe.events
Routing Key: webhook.received
Queue Bindings: gobe.system.events, gobe.mcp.tasks

AMQP Message Format:

{
  "id": "uuid",
  "source": "github",
  "event_type": "push",
  "payload": {...},
  "headers": {...},
  "timestamp": "2025-01-20T12:00:00Z",
  "processed": false,
  "status": "received"
}

Usage

CLI

Start the main server:

./gobe start -p 3666 -b "0.0.0.0"

This starts the server, generates certificates, sets up databases, and begins listening for requests!

See all available commands:

./gobe --help

Main commands:

Command	Function
`start`	Starts the server
`stop`	Safely stops the server
`restart`	Restarts all services
`status`	Shows the status of the server and active services
`config`	Generates an initial configuration file
`logs`	Displays server logs

Configuration

GoBE can run without any initial configuration, but supports customization via YAML/JSON files. By default, everything is generated automatically on first use.

Example configuration:

port: 3666
bindAddress: 0.0.0.0
database:
  type: postgres
  host: localhost
  port: 5432
  user: gobe
  password: secure

Messaging Integrations

WhatsApp and Telegram bots can be configured via the config/discord_config.json file under the integrations section:

{
  "integrations": {
    "whatsapp": {
      "enabled": true,
      "access_token": "<token>",
      "verify_token": "<verify>",
      "phone_number_id": "<number>",
      "webhook_url": "https://your.server/whatsapp/webhook"
    },
    "telegram": {
      "enabled": true,
      "bot_token": "<bot token>",
      "webhook_url": "https://your.server/telegram/webhook",
      "allowed_updates": ["message", "callback_query"]
    }
  }
}

After setting up the file or environment variables, the server will expose the following endpoints:

POST /api/v1/whatsapp/send and /api/v1/whatsapp/webhook
POST /api/v1/telegram/send and /api/v1/telegram/webhook

Each route also provides a /ping endpoint for health checks.

API Reference

Core Endpoints

Category	Method	Endpoint	Description
Health	`GET`	`/health`	Basic health check
Health	`GET`	`/healthz`	Kubernetes-style health check
Health	`GET`	`/status`	Detailed system status
Health	`GET`	`/api/v1/health`	API health with metrics

AI Provider Endpoints

Method	Endpoint	Description	Streaming
`GET`	`/providers`	List all AI providers and availability	❌
`POST`	`/chat`	Chat with AI providers	✅ SSE
`POST`	`/v1/advise`	Get AI advice/recommendations	✅ SSE

MCP (Model Context Protocol) Endpoints

Method	Endpoint	Description	Auth
`GET`	`/mcp/tools`	List available MCP tools	Bearer
`POST`	`/mcp/exec`	Execute MCP tool	Bearer

Webhook Endpoints

Method	Endpoint	Description	Auth
`POST`	`/v1/webhooks`	Receive webhook events	Bearer
`GET`	`/v1/webhooks/health`	Webhook system health	Bearer
`GET`	`/v1/webhooks/events`	List webhook events (paginated)	Bearer
`GET`	`/v1/webhooks/events/:id`	Get specific webhook event	Bearer
`POST`	`/v1/webhooks/retry`	Retry failed webhook events	Bearer

System Monitoring Endpoints

Method	Endpoint	Description	Auth
`GET`	`/api/v1/mcp/system/info`	System information	Bearer
`GET`	`/api/v1/mcp/system/cpu-info`	CPU metrics	Bearer
`GET`	`/api/v1/mcp/system/memory-info`	Memory metrics	Bearer
`GET`	`/api/v1/mcp/system/disk-info`	Disk metrics	Bearer

Scheduler Endpoints

Method	Endpoint	Description	Auth
`GET`	`/health/scheduler/stats`	Scheduler statistics	Bearer
`POST`	`/health/scheduler/force`	Force scheduler run	Bearer

Web UI Endpoints

Method	Endpoint	Description	Auth
`GET`	`/`	Serve web UI root	Public
`GET`	`/app/*path`	Serve web UI application	Public

Authentication

GoBE uses Bearer Token authentication for protected endpoints:

# Get token (implementation depends on your auth setup)
TOKEN="your-jwt-token-here"

# Use token in requests
curl -H "Authorization: Bearer $TOKEN" http://localhost:3666/mcp/tools

Response Formats

Success Response

{
  "status": "success",
  "data": { ... },
  "timestamp": "2025-01-20T12:00:00Z"
}

Error Response

{
  "status": "error",
  "message": "Error description",
  "code": 400,
  "timestamp": "2025-01-20T12:00:00Z"
}

Streaming Response (SSE)

data: {"content": "Hello", "done": false}

data: {"content": " world!", "done": false}

data: {"done": true, "usage": {"total_tokens": 10, "cost_usd": 0.0002}}

Rate Limiting

Default: 100 requests per minute per IP
AI Endpoints: 30 requests per minute per API key
Headers: X-RateLimit-Limit, X-RateLimit-Remaining, X-RateLimit-Reset

CORS Support

CORS is enabled for web UI integration:

// JavaScript frontend example
fetch('http://localhost:3666/providers')
  .then(response => response.json())
  .then(providers => console.log(providers));

Roadmap

✅ Completed (v1.3.5)

🚧 In Progress (v1.4.0)

Database persistence for webhook events
Extended MCP tool library (file operations, network tools)
WebSocket support for real-time MCP communication
Prometheus integration for monitoring

📋 Planned (v1.5.0+)

Plugin system for external tool registration
Advanced security policies and RBAC
Grafana integration for metrics visualization
Multi-tenant support
Advanced AI workflow orchestration
Custom middleware support

🎯 Next Milestones

v1.4.0 - Database persistence, WebSocket MCP, monitoring
v1.5.0 - Plugin system and advanced security
v2.0.0 - Multi-tenant platform with workflow orchestration

Contributing

Contributions are welcome! Feel free to open issues or submit pull requests. See the Contribution Guide for more details.

Contact

💌 Developer: Rafael Mori 💼 Follow me on GitHub I'm open to collaborations and new ideas. If you found the project interesting, get in touch!

Risks & Mitigations

• Zero-config may hide necessary configurations → Verbose logs + override documentation • MCP registry thread-safety → RWMutex implemented + concurrency tests • Dependency on gdbase for DB → Fallback SQLite always available • Certificate auto-generation → Keyring backup + automatic regeneration

Next Steps

Extend MCP tools - file operations, network diagnostics, database queries
WebSocket MCP - real-time tool communication for AI agents
Plugin system - external tool registration via shared libraries
Advanced monitoring - Prometheus metrics + Grafana dashboards
Security hardening - RBAC, audit logs, policy enforcement

Changelog

v1.3.5 (2025-01-20)

✅ AI Provider Ecosystem: Complete OpenAI, Anthropic, Gemini, Groq integration
✅ Streaming AI Responses: Server-Sent Events with cost tracking
✅ Production Webhooks: Full webhook system with AMQP and retry logic
✅ Enhanced MCP Tools: Added shell.command with security whitelisting
✅ Discord Integration: Real MCP tool execution via Discord bot
✅ Comprehensive API: Complete REST API with detailed documentation

v1.3.4 (2024-12-23)

✅ MCP Protocol implementation with dynamic registry
✅ Built-in system.status tool with runtime metrics
✅ Thread-safe tool execution with comprehensive tests
✅ Documentation updated following Kubex standards
✅ Zero-config MCP endpoints (/mcp/tools, /mcp/exec)

Name		Name	Last commit message	Last commit date
Latest commit History 126 Commits
.codex		.codex
.github		.github
cmd		cmd
config		config
docs		docs
factory		factory
internal		internal
support		support
tests		tests
web		web
.gitignore		.gitignore
.hash		.hash
.hintrc		.hintrc
.nvmrc		.nvmrc
.secrets.baseline		.secrets.baseline
AGENTS.md		AGENTS.md
CODE_OF_CONDUCT.md		CODE_OF_CONDUCT.md
CONTRIBUTING.md		CONTRIBUTING.md
Dockerfile		Dockerfile
LICENSE		LICENSE
Makefile		Makefile
NOTICE.md		NOTICE.md
README.md		README.md
README.pt-BR.md		README.pt-BR.md
SECURITY.md		SECURITY.md
docker-compose.yml		docker-compose.yml
go.mod		go.mod
go.sum		go.sum
gobe.go		gobe.go

License

kubex-ecosystem/gobe

Folders and files

Latest commit

History

Repository files navigation

GoBE - Modular & Secure Back-end

TL;DR

Table of Contents

About the Project

Mission Alignment

Current Status - Production Ready

Project Evolution

Recent Major Additions (v1.3.5)

Features

🤖 AI Integration

🔧 System Integration

🏗️ Core Platform

How to Run

Quick Start

Requirements

Build Options

AI Providers

Available Providers

Usage Examples

List Available Providers

Chat with Streaming

Provider Switching

External API Keys

Provider Configuration

Cost Tracking

MCP Support

Available Endpoints

Built-in Tools

Example Usage

List Available Tools

System Status (Basic)

System Status (Detailed)

Execute Shell Commands

Security Features

MCP Architecture

Webhooks

Features

Available Endpoints

Usage Examples

Receive Webhooks

Check System Health

List Webhook Events

Get Event Details

Retry Failed Events

Webhook Event Types

Integration Examples

GitHub Integration

Discord Bot Integration

AMQP Integration

Usage

CLI

Configuration

Messaging Integrations

API Reference

Core Endpoints

AI Provider Endpoints

MCP (Model Context Protocol) Endpoints

Webhook Endpoints

System Monitoring Endpoints

Scheduler Endpoints

Web UI Endpoints

Authentication

Response Formats

Success Response

Error Response

Streaming Response (SSE)

Rate Limiting

CORS Support

Roadmap

✅ Completed (v1.3.5)

🚧 In Progress (v1.4.0)

📋 Planned (v1.5.0+)

🎯 Next Milestones

Contributing