FlowFuse

FlowForge is a self‑hosted platform that extends Node‑RED with full DevOps tooling, team collaboration, and AI‑powered assistance. From a developer’s standpoint it acts as a middleware orchestration layer that manages the entire lifecycle of Node‑RED flows: source control, continuous integration, environment promotion, and remote deployment to edge or cloud targets. The core idea is to treat a Node‑RED instance as a first‑class artifact that can be versioned, reviewed, and rolled out across multiple environments with minimal manual effort.

Technical Stack & Architecture

• Core Runtime: Node‑RED (Node.js 14+) runs inside a Docker container that FlowForge manages. The platform itself is built on NestJS (TypeScript) for the backend API and a React‑based UI.
• Data Layer: PostgreSQL stores user accounts, project metadata, and deployment manifests. Redis is used for session caching and message queuing.
• Deployment Engine: FlowForge orchestrates deployments via Docker Compose or Kubernetes manifests, generating snapshots of a Node‑RED instance (the node-red.json plus assets) that can be pushed to any target.
• AI Integration: A lightweight microservice proxies calls to OpenAI’s GPT‑4.1 Mini, providing a copilot that can generate flow nodes, suggest code snippets, or explain existing flows.

Core Capabilities & APIs

• Version Control: Projects are stored in a Git repository; FlowForge exposes a REST API to create, pull, and merge branches programmatically.
• CI/CD Pipelines: Custom pipeline definitions (YAML) trigger on push events, run tests against the Node‑RED instance, and promote snapshots to staging/production.
• Remote Deployment: The deploy API accepts a snapshot and target definition (Docker host, Kubernetes cluster) and performs the rollout with health checks.
• Webhooks & Events: Developers can subscribe to flow‑execution events, node status changes, or deployment outcomes.
• Plugin Architecture: The platform supports third‑party plugins written in TypeScript that can extend the UI, add new node types, or hook into deployment workflows.

Deployment & Infrastructure

• Self‑Hosted: A single Docker Compose file deploys the entire stack; for larger scale, a Helm chart can be used on any Kubernetes cluster.
• Scalability: Horizontal scaling is achieved by running multiple FlowForge instances behind a load balancer; PostgreSQL can be set up as a read‑replica cluster.
• Edge Compatibility: Snapshots are lightweight, making them ideal for constrained devices; FlowForge can push to Raspberry Pi or industrial gateways via SSH.

Integration & Extensibility

• Node‑RED Ecosystem: All existing Node‑RED nodes remain usable; FlowForge adds a “publish to FlowForge” button that packages the flow into a deployable snapshot.
• External APIs: OAuth2, LDAP, and SAML are supported for authentication; the platform exposes an OpenAPI spec for full automation.
• Custom Workflows: Users can define custom pre‑deploy scripts (bash, Python) that run inside the snapshot container before deployment.

Developer Experience

• Configuration: A declarative flowforge.yaml lets teams define users, projects, and deployment targets in code.
• Documentation: The official docs include a comprehensive API reference, example CI/CD pipelines, and a migration guide from vanilla Node‑RED.
• Community: The GitHub repo has an active issue tracker and a Slack channel for real‑time support. Regular security audits are published in the SECURITY.md.

Use Cases

1. Industrial IoT – Deploy Node‑RED flows that read Modbus/OPC‑UA sensors, convert data to MQTT, and publish dashboards; FlowForge manages rollout across multiple factories.
2. Edge Computing – Package a flow as a snapshot and push it to hundreds of Raspberry Pi gateways with zero‑touch updates.
3. Rapid Prototyping – Use the AI copilot to auto‑generate node configurations, then immediately test and promote the flow via CI/CD.

Advantages

• Unified DevOps: No separate Git, CI, or deployment tooling; everything is integrated into a single platform.
• Reduced Operational Overhead: Snapshots are idempotent, enabling rollbacks and blue‑green deployments with a single API call.
• Open‑Source Flexibility: Self‑hosted deployment removes vendor lock‑in; licensing is MIT, allowing commercial use without fees.
• AI Acceleration: Built‑in GPT assistance cuts development time by up to 10× for complex data pipelines.

FlowForge turns Node‑RED from a single‑user prototyping tool into a production‑grade, collaborative platform that scales across edge and cloud environments while keeping the low‑code experience intact.