ai-dash-macOS

Universal LLM Control Center for macOS Apple Silicon

A production-grade, local-first AI dashboard supporting Ollama, OpenAI-compatible APIs, and optional Anthropic integration — built exclusively for macOS ARM64.

✨ Features

🧠 Universal LLM Router

Multi-provider support — Ollama (local default), OpenAI-compatible, Anthropic (optional)
Runtime switching — Change providers and models on-the-fly from the dashboard
Adapter pattern — Clean abstraction over provider APIs with unified interface
SSE Streaming — Real-time token streaming with Server-Sent Events

💬 Local ChatGPT Replacement

Full chat interface — Conversation sidebar, message bubbles, streaming text with cursor animation
Agent selector — Choose between chat, summary, and analysis agents
Model selector — Auto-detects available Ollama models with dropdown selection
Stream toggle — Switch between streaming SSE and standard request/response
Conversation management — Create, delete, and switch between conversations
Persistent memory — In-memory conversation store with message history

🤖 Multi-Agent System

3 built-in agents — Chat (general), Summary (text summarization), Analysis (deep analysis)
Agent orchestration — Concurrent execution with configurable limits
Task executor — Async task queue with completion/failure tracking
Custom agents — Extensible registry for adding new agent types

💻 Coding Engine

AI code generation — Generate code in any language via LLM
Code analysis — Static analysis and AI-powered code review
Code execution — Sandboxed code execution with output capture
File management — Browse, read, and manage project files

🔄 Dual Automation System

Custom Workflow Engine — 6 step types (llm_query, agent_task, code_execute, http_request, condition, transform), sequential execution with context passing, conditional branching
n8n Connector — Full integration with n8n workflow automation:
- Connection settings UI (URL, API key, enable/disable)
- List, activate, deactivate, and execute n8n workflows
- Webhook triggering
- Execution history viewer
- Runtime configuration without restart
Scheduler — Interval-based recurring job execution with max runs

📊 Unified Dashboard

React 19 + TypeScript + Vite — Modern, fast frontend
Tailwind CSS — Ultra-modern dark theme UI
Real-time polling — Auto-refreshing system metrics
Multi-page layout — Overview, Chat, LLM, Agents, Coding, Automation pages
Responsive design — Clean layout on all screen sizes

🍎 Apple Silicon Optimized

ARM64 native — No Rosetta, optimized for unified memory architecture
Memory-aware — Intelligent model gating based on available RAM (configurable threshold)
System monitoring — Real-time CPU, memory, swap, and GPU metrics
Auto-detection — Platform validation at startup

📋 Requirements

Requirement	Version
macOS	Apple Silicon (M1/M2/M3/M4)
Python	3.11+
Node.js	18+ (ARM64)
Ollama	Latest (install)
RAM	16GB+ unified memory (28GB recommended)

🚀 Quick Start

# Clone
git clone https://github.com/edlaracuente/ai-dash-macOS.git
cd ai-dash-macOS

# Create virtual environment
python3 -m venv venv
source venv/bin/activate

# Install backend dependencies
pip install -r requirements.txt

# Install frontend dependencies
cd frontend && npm install && cd ..

# Copy environment config
cp .env.example .env

# Make sure Ollama is running with a model
ollama pull llama3.1:latest
ollama serve  # if not already running

# Start backend (terminal 1)
uvicorn backend.app.main:app --host 127.0.0.1 --port 8000 --reload

# Start frontend (terminal 2)
cd frontend && npm run dev

# Open dashboard
open http://localhost:3000

Health Check

curl http://localhost:8000/health | jq

🏗 Project Structure

ai-dash-macOS/
├── backend/
│   └── app/
│       ├── api/              # REST endpoints
│       │   ├── health.py     # Health check & system info
│       │   ├── llm.py        # LLM config, models, status
│       │   ├── agents.py     # Agent list, task execution
│       │   ├── conversations.py  # Chat, messages, streaming
│       │   ├── automation.py # Workflows, scheduler, n8n
│       │   └── coding.py     # File management, code ops
│       ├── core/             # Config, validation, memory monitor
│       │   ├── config.py     # Pydantic Settings (.env)
│       │   ├── validation.py # ARM64 & Ollama checks
│       │   └── memory.py     # Memory monitoring & gating
│       ├── llm/              # LLM provider system
│       │   ├── router.py     # Multi-provider router
│       │   ├── adapters/     # Ollama, OpenAI, Anthropic
│       │   └── streaming.py  # SSE streaming
│       ├── agents/           # Agent system
│       │   ├── registry.py   # Agent registration
│       │   ├── executor.py   # Async task executor
│       │   └── builtins.py   # Chat, Summary, Analysis agents
│       ├── automation/       # Automation system
│       │   ├── engine.py     # Custom workflow engine
│       │   ├── scheduler.py  # Interval-based scheduler
│       │   ├── n8n.py        # n8n connector service
│       │   └── models.py     # Workflow/step models
│       ├── coding/           # Coding engine
│       │   ├── generator.py  # AI code generation
│       │   ├── analyzer.py   # Code analysis
│       │   ├── executor.py   # Sandboxed execution
│       │   └── files.py      # File management
│       ├── memory/           # Memory & knowledge
│       │   └── store.py      # Conversation store
│       └── monitoring/       # System monitoring
│           └── system.py     # CPU, RAM, GPU metrics
├── frontend/
│   └── src/
│       ├── App.tsx           # Router & layout
│       ├── api.ts            # API client (all endpoints)
│       ├── types.ts          # TypeScript interfaces
│       ├── components/       # Reusable UI components
│       │   ├── Layout.tsx    # Sidebar navigation
│       │   └── Card.tsx      # Card, Stat, Badge components
│       ├── hooks/
│       │   └── usePoll.ts    # Auto-polling hook
│       └── pages/
│           ├── OverviewPage.tsx   # System dashboard
│           ├── ChatPage.tsx       # Chat interface
│           ├── LLMPage.tsx        # LLM management
│           ├── AgentsPage.tsx     # Agent management
│           ├── CodingPage.tsx     # Coding tools
│           └── AutomationPage.tsx # Workflows + n8n
├── docs/               # Phase documentation
├── scripts/            # Utility scripts
├── docker/             # Docker configs (ARM64)
├── .github/workflows/  # CI pipeline
├── .env.example        # Environment template
├── requirements.txt    # Python dependencies
└── package.json        # Root package.json

🔌 API Endpoints

Health

Method	Endpoint	Description
GET	`/health`	System health, platform, memory, Ollama status

LLM

Method	Endpoint	Description
GET	`/api/llm/status`	Router status, active provider, model counts
GET	`/api/llm/config`	Current LLM configuration
PUT	`/api/llm/config`	Update provider/model at runtime
GET	`/api/llm/models`	List all available models

Chat / Conversations

Method	Endpoint	Description
GET	`/api/conversations/`	List all conversations
POST	`/api/conversations/`	Create new conversation
GET	`/api/conversations/{id}`	Get conversation with messages
DELETE	`/api/conversations/{id}`	Delete conversation
GET	`/api/conversations/{id}/messages`	Get message history
POST	`/api/conversations/{id}/chat`	Send message (supports SSE streaming)
GET	`/api/conversations/stats`	Memory usage statistics

Agents

Method	Endpoint	Description
GET	`/api/agents/list`	List registered agents
GET	`/api/agents/status`	Executor status & metrics

Automation

Method	Endpoint	Description
GET	`/api/automation/workflows`	List custom workflows
GET	`/api/automation/scheduler/status`	Scheduler status

n8n Integration

Method	Endpoint	Description
GET	`/api/automation/n8n/status`	n8n connection status
GET	`/api/automation/n8n/workflows`	List n8n workflows
POST	`/api/automation/n8n/workflows/{id}/activate`	Activate workflow
POST	`/api/automation/n8n/workflows/{id}/deactivate`	Deactivate workflow
POST	`/api/automation/n8n/workflows/{id}/execute`	Execute workflow
POST	`/api/automation/n8n/webhook/{path}`	Trigger webhook
GET	`/api/automation/n8n/executions`	Execution history
PUT	`/api/automation/n8n/config`	Update n8n config at runtime

Coding

Method	Endpoint	Description
GET	`/api/coding/files`	List project files

⚙️ Environment Variables

All configuration is managed via .env. See .env.example for the complete list.

Variable	Default	Description
`OLLAMA_URL`	`http://localhost:11434`	Ollama API URL
`OLLAMA_MODEL`	`llama3.1:latest`	Default Ollama model
`DEFAULT_PROVIDER`	`ollama`	Active LLM provider
`ALLOW_REMOTE_PROVIDERS`	`false`	Allow OpenAI/Anthropic
`OPENAI_API_KEY`	—	OpenAI API key (optional)
`ANTHROPIC_API_KEY`	—	Anthropic API key (optional)
`N8N_URL`	`http://localhost:5678`	n8n instance URL
`N8N_API_KEY`	—	n8n API key
`N8N_ENABLED`	`false`	Enable n8n integration
`MEMORY_THRESHOLD_PERCENT`	`90`	Memory gating threshold
`MAX_HEAVY_MODELS`	`2`	Concurrent model limit
`MAX_AGENT_CONCURRENCY`	`3`	Parallel agent limit

🧪 Testing

# Run all tests
pytest backend/tests/ -v

# Run with coverage
pytest backend/tests/ --cov=backend/app -v

🐳 Docker (ARM64)

docker-compose up --build

📖 Documentation

Document	Description
ARCHITECTURE.md	System architecture & design decisions
ENVIRONMENT.md	Environment setup guide
PHASE_0.md	Local Validation — ARM64, Ollama, health
PHASE_1.md	LLM Router Core — Adapters, streaming
PHASE_2.md	Agents + Memory — Registry, executor, store
PHASE_3.md	Coding Engine — Generator, analyzer, executor
PHASE_4.md	Automation — Workflows, scheduler
PHASE_5.md	Unified Dashboard — React UI
CONTRIBUTING.md	Contribution guidelines

📦 Build Phases

Phase	Description	Status
0	Local Validation — ARM64, memory monitoring, Ollama check	✅
1	LLM Router Core — Provider adapters, SSE streaming, 130 tests	✅
2	Agents + Memory — Registry, executor, conversation store	✅
3	Coding Engine — Generator, analyzer, executor, file manager	✅
4	Automation — Workflow engine, scheduler, n8n connector	✅
5	Unified Dashboard — React 19, Chat UI, real-time monitoring	✅

📸 Screenshots

Dashboard Overview

System health, LLM status, agent metrics, and conversation stats at a glance.

Chat Interface

Full ChatGPT-like experience with streaming responses, agent selection, and model switching.

Automation

Dual automation system with custom workflow engine and n8n connector integration.

🛠 Tech Stack

Layer	Technology
Backend	Python 3.11+, FastAPI, Pydantic v2, uvicorn, httpx
Frontend	React 19, TypeScript, Vite 6, Tailwind CSS
LLM	Ollama (local), OpenAI API, Anthropic API
Automation	Custom engine + n8n integration
Database	In-memory (conversation store)
Streaming	Server-Sent Events (SSE)
Testing	pytest, 170+ tests
CI/CD	GitHub Actions

📄 License

MIT

Built with ❤️ for Apple Silicon by Ed Laracuente

Name		Name	Last commit message	Last commit date
Latest commit History 14 Commits
.github/workflows		.github/workflows
backend		backend
docker		docker
docs		docs
frontend		frontend
scripts		scripts
.env.example		.env.example
.gitignore		.gitignore
CONTRIBUTING.md		CONTRIBUTING.md
LICENSE		LICENSE
README.md		README.md
docker-compose.yml		docker-compose.yml
package.json		package.json
pytest.ini		pytest.ini
requirements.txt		requirements.txt

Folders and files

Latest commit

History

Repository files navigation

ai-dash-macOS

✨ Features

🧠 Universal LLM Router

💬 Local ChatGPT Replacement

🤖 Multi-Agent System

💻 Coding Engine

🔄 Dual Automation System

📊 Unified Dashboard

🍎 Apple Silicon Optimized

📋 Requirements

🚀 Quick Start

Health Check

🏗 Project Structure

🔌 API Endpoints

Health

LLM

Chat / Conversations

Agents

Automation

n8n Integration

Coding

⚙️ Environment Variables

🧪 Testing

🐳 Docker (ARM64)

📖 Documentation

📦 Build Phases

📸 Screenshots

Dashboard Overview

Chat Interface

Automation

🛠 Tech Stack

📄 License

About

Topics

Resources

License

Contributing

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages