AI Content OS

Multi-agent AI content creation platform with visual workflows, 8 specialized agents, and real-time collaboration

---

AI Content OS is an open-source, multi-agent content creation platform that orchestrates 8 specialized AI agents through visual workflows to research, write, optimize, translate, design, and publish content — all with brand consistency, cost control, and real-time streaming.

---

Features

Feature	Description
8 AI Agents	Research, Writer, SEO, Editor, Translator, Designer, Publisher, Reviewer — each a specialist
Visual Workflow Builder	Drag-and-drop pipelines with conditions, retries, parallelism, and human approval gates
Multi-Model Routing	Intelligent selection across GPT-4o, Claude, and DeepSeek based on task, cost, and quality
Real-Time Streaming Chat	WebSocket-powered chat with streaming LLM responses and agent collaboration
Brand Consistency Engine	Define voice, tone, style guides, and vocabulary — enforced across all content
RAG-Powered Research Studio	Upload documents, build knowledge bases, and ground AI responses in your data
Background Task Queue	Async job processing for long-running workflows and batch operations
WebSocket Real-Time Updates	Live progress, status, and streaming output to any connected client
Plugin System	Extend agents, tools, workflows, and integrations with a first-class SDK
Admin Dashboard	User management, system health, cost analytics, and audit logs
Dark Mode	Full dark mode support in the React frontend

---

Architecture

flowchart TD
    FE["Frontend<br/>React 18 / TypeScript / Tailwind / Vite<br/>Dashboard - Workflow Builder - Chat - Editor"]
    FE -->|"REST + WebSocket"| API

    API["API Layer - FastAPI<br/>JWT Auth / Rate Limiting / CORS<br/>/auth /chat /agents /content /workflows /brands"]

    subgraph Core["Core"]
        SVC["Services<br/>Content / Brand / Memory<br/>Model Router / Analytics"]
        AGENTS["8 Agents<br/>Research - Writer - SEO - Editor<br/>Translator - Designer - Publisher - Reviewer"]
        WF["Workflow Engine<br/>Conditions / Retries / Approval Gates"]
        PLUG["Plugins<br/>Agent / Tool / Integration"]
    end
    API --> SVC
    API --> AGENTS
    API --> WF
    API --> PLUG
    WF --> AGENTS
    SVC --> AGENTS

    MODEL["Model Layer<br/>OpenAI GPT-4o - Anthropic Claude - DeepSeek<br/>cost optimization / quality routing / fallback"]
    AGENTS --> MODEL

    DATA["Data and Infrastructure<br/>SQLite-PostgreSQL / SQLAlchemy / Redis / Celery<br/>Alembic - Docker - GitHub Actions"]
    SVC --> DATA
    WF --> DATA

Loading

---

Quick Start

With Docker (recommended)

git clone https://github.com/WAHIB-EL-KHADIRI/ai_content_factory.git
cd ai-content-os
cp .env.example .env
# Edit .env with your API keys
docker compose up -d
# Open http://localhost:3000 (frontend) or http://localhost:8000/docs (API)

Manual Setup

git clone https://github.com/WAHIB-EL-KHADIRI/ai_content_factory.git
cd ai-content-os

# Backend
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt
cp .env.example .env

# Start backend
python -m backend.api.app

# Frontend (separate terminal)
cd frontend
npm install
npm run dev

Verify

curl http://localhost:8000/health
# {"status":"healthy","version":"1.0.0"}

---

API Reference

Authentication

# Register
curl -X POST http://localhost:8000/api/v1/auth/register \
  -H "Content-Type: application/json" \
  -d '{"username":"demo","email":"demo@example.com","password":"securepass123"}'

# Login
curl -X POST http://localhost:8000/api/v1/auth/login \
  -H "Content-Type: application/json" \
  -d '{"username":"demo","password":"securepass123"}'

Chat (Streaming)

curl -X POST http://localhost:8000/api/v1/chat/stream \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <token>" \
  -d '{"message":"Write a blog post about AI trends in 2026","model":"gpt-4o"}'

Content Creation

curl -X POST http://localhost:8000/api/v1/content \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <token>" \
  -d '{
    "project_id": "my-project",
    "topic": "The Future of AI in Healthcare",
    "content_type": "article",
    "word_count": 1500
  }'

Agent Execution

curl -X POST http://localhost:8000/api/v1/agents/run \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <token>" \
  -d '{"agent_role":"seo","task":{"action":"keywords","topic":"AI healthcare"}}'

Workflow Management

# Create from template
curl -X POST "http://localhost:8000/api/v1/workflows/templates/article?topic=AI+in+Healthcare" \
  -H "Authorization: Bearer <token>"

# Execute workflow
curl -X POST http://localhost:8000/api/v1/workflows/run \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <token>" \
  -d '{"workflow_id":"<id>","context":{"project_id":"my-project"}}'

Brand Management

curl -X POST http://localhost:8000/api/v1/brands \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer <token>" \
  -d '{"name":"Acme Corp","voice":"professional","tone":"friendly","style_guide":"Use simple language"}'

Full interactive API docs at http://localhost:8000/docs (Swagger) and http://localhost:8000/redoc (ReDoc).

---

Agents

Agent	Role	What It Does
Research	research	Gathers information, analyzes topics, and synthesizes findings
Writer	writer	Creates articles, scripts, social posts, and long-form content
SEO	seo	Keyword research, on-page optimization, meta tag generation
Editor	editor	Grammar, style, clarity, tone consistency, and restructuring
Translator	translator	Multi-language translation with cultural localization
Designer	designer	Visual content concepts, image prompt generation, layout suggestions
Publisher	publisher	Platform-specific formatting, scheduling, and distribution
Reviewer	reviewer	Quality assurance, fact-checking, compliance, and final approval

---

Tech Stack

Layer	Technology	Purpose
Backend	FastAPI, SQLAlchemy, Celery-compatible	Async REST API with ORM and task queue
Frontend	React 18, TypeScript, Tailwind CSS, Vite	Modern SPA with type safety and fast builds
AI Models	OpenAI, Anthropic, DeepSeek	Multi-provider model routing and fallbacks
Database	SQLite (dev) / PostgreSQL (prod)	Relational storage with Alembic migrations
Cache/Queue	Redis	Background job processing and caching
Infrastructure	Docker, Docker Compose	Containerized development and deployment
CI/CD	GitHub Actions	Automated testing, linting, and deployment
Code Quality	Ruff, MyPy, Pytest	Linting, type checking, and testing

---

Project Structure

ai-content-os/
├── backend/
│   ├── api/                # FastAPI routes, middleware, WebSocket
│   ├── core/               # Configuration, auth, security
│   ├── db/                 # SQLAlchemy models and repositories
│   ├── agents/             # 8 specialized AI agents
│   ├── services/           # Business logic (content, brand, memory)
│   ├── workflows/          # Workflow engine, builder, conditions
│   ├── analytics/          # Cost tracking, usage metrics
│   └── plugins/            # Plugin SDK and registry
├── frontend/
│   ├── src/
│   │   ├── components/     # React UI components
│   │   ├── pages/          # Route pages
│   │   └── hooks/          # Custom React hooks
│   ├── package.json
│   └── vite.config.ts
├── tests/
│   ├── unit/               # Unit tests
│   ├── integration/        # Integration tests
│   └── e2e/                # End-to-end tests
├── docker/                 # Docker configuration
├── .github/workflows/      # CI/CD pipelines
├── docs/                   # Documentation
├── alembic/                # Database migrations
├── docker-compose.yml
├── requirements.txt
└── .env.example

---

Configuration

All configuration is managed via environment variables. Copy .env.example to .env and set your keys:

Variable	Description	Required
OPENAI_API_KEY	OpenAI API key	Yes
ANTHROPIC_API_KEY	Anthropic API key	No
DEEPSEEK_API_KEY	DeepSeek API key	No
DATABASE_URL	Database connection string	Yes
SECRET_KEY	Application secret for JWT	Yes
REDIS_URL	Redis connection for task queue	No

---

Testing

Last measured in CI on 2026-09-16: 245 passed, 59% line coverage (1,302 of 3,207 statements uncovered).

That figure is quoted from the CI log and dated, rather than shown as a badge, because the Codecov upload is not currently authenticated — a badge would either sit stale or read unknown. Re-run the command below to check it yourself; if the date above is old, trust the command and not the number.

# Unit tests
pytest tests/unit/ -v

# Integration tests
pytest tests/integration/ -v

# All tests with coverage
pytest tests/ -v --cov=backend

# Lint and type check
ruff check backend/
mypy backend/

---

Contributing

See CONTRIBUTING.md for development setup, code style, and PR guidelines.

---

License

MIT License — see LICENSE for details.

---

Built and maintained by WAHIB EL KHADIRI. Bugs, feature requests and questions all go to Issues.