Skip to content
BranchesTags

Repository files navigation

OmniDev AI Platform

OmniDev AI Platform

All-in-One AI Development Platform

Open your browser, write code with AI, manage projects, and deploy — all in one place.

中文文档 · Quickstart · Architecture · Docs

What is OmniDev?

OmniDev integrates an AI coding assistant, online IDE, Agent orchestration, RAG knowledge base, CI/CD pipelines, and monitoring into a single platform. It's designed for developers and teams who want to build, ship, and operate software without switching between a dozen tools.

Key Features

Online IDE

  • Monaco Editor — syntax highlighting, multi-cursor, minimap, file search & replace
  • AI Code Completion — Tab completion, multi-line suggestions, context-aware via LSP
  • Agent Code Editing — AI Agent directly modifies code with Diff preview, accept/reject/undo
  • Terminal Emulator — xterm.js, multi-tab, SSH, PTY support
  • Git Integration — commit/push/pull/branch/merge, visual Diff
  • Code Review — inline comments, AI-assisted review

Agent System

  • Natural Language Tasks — describe what you want, Agent decomposes, plans, and executes
  • Tool Calling — built-in tools + custom MCP tools with Function Calling support
  • Visual Orchestration — drag-and-drop canvas, node wiring, conditional branching
  • Sandbox Execution — Docker isolation, resource limits, filesystem mounting
  • Execution Monitoring — real-time logs, step tracing, intermediate state inspection
  • Agent Templates — pre-built: code review, test generation, doc generation, etc.
  • Multi-Agent Collaboration — CrewAI-style, role assignment, task delegation
  • Agent Marketplace — export configs, ratings, install stats

RAG Knowledge Base

  • Multi-format upload (PDF, Markdown, code files)
  • Hybrid search: vector (pgvector) + keyword (Elasticsearch)
  • Code-aware chunking & indexing
  • Integration with Chat and Agent for context injection

CI/CD & Deployment

  • Natural language pipeline configuration
  • Kubernetes-native deployment
  • Multi-cloud support (AWS/GCP/Azure)
  • Git-triggered automated pipelines

Monitoring & Observability

  • Built-in OpenTelemetry integration
  • Prometheus + Grafana + Loki + Jaeger stack
  • AI-powered anomaly detection
  • Custom alerting rules

Tech Stack

Layer Technology
Backend Go 1.22+, Gin, gRPC, Wire (DI)
Frontend Next.js 15 (App Router), React 19, Tailwind CSS, Shadcn/ui
Database PostgreSQL 16 + pgvector, Redis 7, Kafka (KRaft)
Storage MinIO (S3 compatible)
Search Elasticsearch 8
Workflow Temporal
Infra Docker, Kubernetes, Helm, Terraform
Observability Prometheus, Grafana, Loki, Jaeger, OpenTelemetry

Architecture

┌─────────────────────────────────────────────────────────────┐
│                    Gateway (Kong/Custom)                      │
│              Rate Limit · Auth · Routing · WAF               │
└──────────┬──────────┬──────────┬──────────┬─────────────────┘
           │          │          │          │
     ┌─────┴───┐ ┌───┴────┐ ┌──┴───┐ ┌───┴─────┐
     │   IDE   │ │  Chat  │ │Agent │ │   RAG   │
     │ Service │ │ Engine │ │System│ │ Service │
     └─────────┘ └────────┘ └──────┘ └─────────┘
           │          │          │          │
     ┌─────┴───┐ ┌───┴────┐ ┌──┴───┐ ┌───┴─────┐
     │  Git    │ │  MCP   │ │Sandbox│ │  MCP   │
     │ Service │ │ Server │ │System │ │ Server │
     └─────────┘ └────────┘ └───────┘ └─────────┘
           │          │          │          │
     ┌─────┴──────────┴──────────┴──────────┴─────┐
     │              Data Layer                      │
     │  PostgreSQL · Redis · Kafka · MinIO · ES    │
     └─────────────────────────────────────────────┘

Project Structure

omnidev-ai-platform/
├── apps/                    # Application services
│   ├── web/                 # Next.js frontend
│   ├── gateway/             # API Gateway (Go/Gin)
│   ├── user/                # User Service (Go)
│   ├── project/             # Project Service (Go)
│   ├── workspace/           # Workspace Service (Go)
│   ├── terminal/            # Terminal Service (Go)
│   ├── file/                # File Service (Go)
│   ├── git/                 # Git Service (Go)
│   ├── chat/                # Chat Service (Go)
│   ├── agent/               # Agent Service (Go)
│   ├── rag/                 # RAG Service (Go)
│   ├── workflow/            # Workflow Service (Go)
│   ├── sandbox/             # Sandbox Service (Go)
│   ├── mcp/                 # MCP Service (Go)
│   ├── billing/             # Billing Service (Go)
│   ├── deploy/              # Deploy Service (Go)
│   └── monitor/             # Monitor Service (Go)
├── packages/                # Shared libraries
│   ├── ui/                  # UI component library
│   ├── utils/               # Common utilities
│   ├── proto/               # gRPC proto definitions
│   └── types/               # Shared TypeScript types
├── tools/                   # Dev tools & scripts
├── deploy/                  # Deployment configs
├── docs/                    # Documentation
└── web-extensions/          # Browser extensions

Quickstart

Prerequisites

  • Go 1.22+
  • Node.js 20+
  • Docker & Docker Compose
  • PostgreSQL 16
  • Redis 7

Run Locally

# Clone the repository
git clone https://github.com/your-org/omnidev-ai-platform.git
cd omnidev-ai-platform

# Start infrastructure
docker-compose up -d postgres redis kafka minio elasticsearch

# Run database migrations
cd tools/migrations && make migrate-up

# Start backend services
cd apps/gateway && go run cmd/main.go

# Start frontend
cd apps/web && pnpm install && pnpm dev

Environment Variables

Copy .env.example to .env and configure:

cp .env.example .env

Key variables:

  • DATABASE_URL — PostgreSQL connection string
  • REDIS_URL — Redis connection string
  • JWT_SECRET — JWT signing secret
  • OPENAI_API_KEY — OpenAI API key (or other provider)

RAG Embedding Configuration

Configure the embedding provider for RAG knowledge base. Provider-based selection instead of model name guessing:

Variable Description Default
RAG_EMBEDDING_PROVIDER Provider name: openai / gemini / ollama openai
RAG_EMBEDDING_MODEL Model identifier (provider-specific) text-embedding-3-small
RAG_EMBEDDING_API_KEY API key override (optional, falls back to provider key) —
RAG_EMBEDDING_BASE_URL Base URL for self-hosted services (Ollama, vLLM, etc.) —

Example configurations:

# OpenAI (default)
RAG_EMBEDDING_PROVIDER=openai
RAG_EMBEDDING_MODEL=text-embedding-3-small

# Gemini
RAG_EMBEDDING_PROVIDER=gemini
RAG_EMBEDDING_MODEL=gemini-embedding-2

# Ollama (local)
RAG_EMBEDDING_PROVIDER=ollama
RAG_EMBEDDING_MODEL=nomic-embed-text
RAG_EMBEDDING_BASE_URL=http://localhost:11434/v1

# Self-hosted OpenAI-compatible (vLLM, text-embedding-inference, etc.)
RAG_EMBEDDING_PROVIDER=openai
RAG_EMBEDDING_MODEL=BAAI/bge-large-zh-v1.5
RAG_EMBEDDING_BASE_URL=http://your-server:8080/v1

To add a new embedding provider, register a factory in embedder/registry.go — zero main.go changes needed.

Competitive Landscape

Capability Competitors OmniDev Differentiator
AI Coding Cursor, GitHub Copilot Multi-model + local models + RAG
Online IDE Codespaces, Gitpod Embedded AI Agent + one-click deploy
Agent Platform OpenHands, AutoGPT Visual orchestration + sandbox + marketplace
RAG / Knowledge ChatGPT Projects, Notion AI Multi-format + hybrid search + code index
CI/CD Jenkins, GitHub Actions Natural language config + K8s native
Deployment Vercel, Railway Multi-cloud + K8s + cost optimization
Monitoring Grafana Cloud Built-in OTel + AI anomaly detection

Target Users

Persona Use Case
Solo Developer Rapid prototype → AI-assisted coding → one-click deploy
Small Team (2-10) Collaborative dev + Agent automation + shared knowledge base
Medium/Large Team (10-100) RBAC + audit + compliance + private deployment
AI App Developer Build/debug/deploy Agents and MCP Servers

Development Roadmap

Milestone Timeline Deliverables
M0 — Foundation Month 1-2 Infrastructure, DB, auth, CI/CD
M1 — Alpha Month 3-4 IDE (Monaco + terminal) + Agent (basic execution + tool calling + sandbox)
M2 — Beta Month 5-6 IDE (Git + code review) + Agent (frontend + monitoring)
M3 — Ecosystem Month 7-9 Agent visual orchestration + marketplace + multi-agent
M4 — GA Month 10-12 Performance + security hardening + plugin system

Documentation

Document Description
Executive Summary Project vision & positioning
Requirements Analysis Functional & non-functional requirements
Feature Boundary Module boundaries & dependencies
Non-Functional Requirements Performance, security, availability
Technology Selection Tech stack comparison & rationale
System Architecture Architecture diagrams (Mermaid)
Database Design ER diagrams & schema
Directory Structure Project layout
Development Standards Coding standards & workflow
Milestone Plan Detailed milestone schedule

Feature Documentation

Document Description
File Upload & Multimodal File upload, document parsing (Tika), multimodal messages

License

MIT

About

No description, website, or topics provided.

Resources

Readme
Activity

Stars

1 star

Watchers

0 watching

Forks

0 forks
Report repository

Releases

Packages

Contributors

Languages