📌Overview

Purpose:
FLUJO aims to provide an open-source, unified platform for orchestrating AI workflows, integrating Model-Context-Protocol (MCP) servers, and seamlessly managing AI tools, all within a user-friendly local environment.

Overview:
FLUJO bridges workflow orchestration with powerful AI integrations, delivering a centralized solution for managing AI models, MCP servers, and complex data flows. Built on the PocketFlowFramework and CLine, FLUJO empowers users to design, configure, and interact with advanced workflows enhanced by AI and external tool integrations—entirely open-source and locally hosted for full control.

Key Features:

Secure Environment & API Key Management:
Store and encrypt API keys and environment variables, enabling global, centralized, and secure access across all application components.
Versatile Model Management:
Effortlessly configure and run multiple AI models (cloud-based or local), customize system prompts, and connect to various providers (like OpenAI or Anthropic).
Comprehensive MCP Server Integration:
Install and manage MCP servers (from GitHub, local, or Docker), inspect tools, and bind environment variables, streamlining model and tool orchestration.
Visual Workflow Orchestration:
Design sophisticated workflows using an interactive visual builder, integrating models, tools, and custom prompts at multiple levels.
Integrated Chat Interface:
Interact with configured flows using a chat UI, supporting rich features like message management, file attachments, and basic transcription.
External Tool Integration:
Provide an OpenAI-compatible endpoint, enabling FLUJO as a backend for other AI applications and developer tools.

FLUJO

Disclaimer

FLUJO is an early preview—please provide feedback and report any issues via GitHub Issues or Discord.

Security Note:
FLUJO currently has extensive logging enabled by default, which may expose your encrypted API keys in terminal output. Avoid sharing terminal output publicly until this is resolved.

Overview

FLUJO is an open-source platform for workflow orchestration, Model-Context-Protocol (MCP) integration, and AI tool management. It provides a unified, local, and open-source interface for managing AI models, MCP servers, and complex workflows.

Features

Environment & API Key Management

Secure storage for environment variables and API keys (encrypted)
Global access to stored credentials throughout the app

Model Management

Configure and use multiple AI models simultaneously
Custom system instructions per model
Integrate with providers like OpenAI, Anthropic, and local models via Ollama

MCP Server Integration

Install MCP servers from GitHub or local filesystem
Manage servers and tools from a unified interface
Connect server environment variables to global storage
Docker support for managing MCP servers

Workflow Orchestration

Visual builder for complex workflows
Connect AI models and manage tool access through flows
Configure prompts at multiple levels (model, flow, node)

Chat Interface

Interact with flows via chat
Edit and manage conversation context
Attach files or audio for LLM processing and transcription

External Tool Integration

OpenAI-compatible endpoint for integration with tools like CLine or Roo
Use FLUJO as a backend for other AI applications

Getting Started

Prerequisites

Node.js (v18 or higher)
npm or yarn

Installation

Clone the repository:

git clone https://github.com/mario-andreschak/FLUJO.git
cd FLUJO

Install dependencies:
```
npm install
# or
yarn install
```
Start the development server:
```
npm run dev
# or
yarn dev
```
Open your browser at http://localhost:4200

Production Build

npm run build
npm start

Desktop Application

npm run electron-dev    # For development
npm run electron-dist   # Build and package

Usage

API Keys

Go to Settings and securely store your API keys.

Models

Models page → "Add Model" → Configure name, provider, key, prompt → Save.

MCP Servers

MCP page → "Add Server" → Install from GitHub or local filesystem → Configure and manage server.

Using SSE MCP-Servers

MCP page → "Add Server" → Select "Local Server" → Set root path to / and configure as needed.

Official Reference Servers

MCP page → "Add Server" → Reference Servers Tab → Refresh (first time), then select and save a server.

Docker-Based MCP Servers

MCP page → "Add Server" → Choose "Docker" → Enter image name and environment variables.

Workflows

Flows page → "Create Flow" → Add/connect nodes → Configure models/tools → Save.

Branching, Loops, Orchestration

Create branches by connecting nodes, define behavior in prompts and node settings.

Chat Interface

Chat page → Select a flow → Interact with the workflow.

MCP Integration

Install/manage MCP servers, inspect tools, link servers to flows, bind server environment variables to global storage.

Docker Installation

Prerequisites

Docker and Docker Compose

Quick Start

Clone the repository:

git clone https://github.com/mario-andreschak/FLUJO.git
cd FLUJO

Start with Docker Compose:
```
docker-compose up -d
```
Use via http://localhost:4200

Advanced Docker

./scripts/build-docker.sh — Build image
./scripts/run-docker.sh — Run container

Options:

--tag=<tag>: Specify image tag
--detached
--no-privileged
--port=<port> (Default: 4200)

License

FLUJO is licensed under the MIT License.

Roadmap

Real-time voice via Whisper.js/OpenWhisper
Visual debugger
MCP Roots: Checkpoints and Restore
Custom MCP prompts
MCP Proxying: Use managed MCP Servers in other clients
Tool integrations: Windsurf, Cursor, CLine
Advanced orchestration: agent-driven, batch jobs
Online template repository for sharing flows/packages
Edge device optimization

Contributing

Fork the repository
Create a feature branch
Commit your changes
Push and open a Pull Request

Contact

GitHub: mario-andreschak
LinkedIn: https://www.linkedin.com/in/mario-andreschak-674033299/

FLUJO - Empowering your AI workflows with open-source orchestration.