Build Your Own
Data Pipeline

Design with Connectors that gather data from any brand of device, transfer them through On-Prem and Cloud MQTT Brokers and feed them into all sorts of Data Applications.

corefluxlogo

Your AI Assistant
Already Knows LoT

Coreflux connects to the AI tools you already use, like Claude, GitHub Copilot, Cursor, and more. Describe what you need in plain language and get accurate, working LoT code grounded in the official documentation.

No guesses, no hallucinated syntax. The real thing.
Connect Your AI Assistant
Try it in the Docs

Most AI Assistants allucinate Industrial Code.

Without access to real documentation, even the best AI models pattern-match their way to code that almost works. Then, it fails on the broker.

Coreflux connects your AI assistant directly to the official docs via MCP.
Every answer is grounded in verified syntax, not training data guesses.

Give Your AI a Library Card for Coreflux.

The Coreflux MCP Server connects your AI assistant directly to the official Coreflux documentation using the Model Context Protocol (an open standard created by Anthropic that lets AI tools call external services during a conversation).

When you ask it a Coreflux question, it searches the documentation, reads the relevant pages, and responds with syntax and examples verified against the actual docs.

AI That Works Where You Work.

Coreflux AI features run across every part of your workflow: in your editor,in your browser, and inside the Hub itself.

The result: accurate LoT code, correct broker configuration and real answers, all inside the platform itself.
VS Code
Write LoT, run LoT, and get AI help, all in one window.
The LoT Notebooks extension for VS Code gives you a Jupyter-style environment
for writing and deploying LoT code. Connect GitHub Copilot with the Coreflux
MCP and your AI assistant can generate, validate, and improve your LoT code
directly inside the notebook, referencing live documentation as it goes.
Write it, run it, watch it execute. AI co-pilot included.
Cursor
Describe what you want. Get working LoT code.
Cursor's native MCP support makes it one of the fastest ways to build with
Coreflux. Add the Coreflux MCP server with one config entry and Cursor's
AI will search the documentation and generate LoT code that actually runs.
Ask it to build an entire automation system. It will consult the docs,
write the Actions, and structure the Models, all while you describe the logic
in plain English.
Claude
Industrial automation logic, described in conversation.
Connect the Coreflux MCP to Claude Desktop or Claude.ai and you have a
documentation-grounded assistant that can explain LoT concepts, generate
working code, review your existing logic, and answer architecture questions with source references you can verify.
Ask Claude to build. Ask it to explain. Ask it to check your work.
It will look up the answer instead of inventing one.
Docs AI Assistant
Ask the documentation a question. Get a direct answer.
Built into docs.coreflux.org, the Docs AI Assistant is trained specifically
on Coreflux documentation. Ask it how to configure a PostgreSQL Route, how
Actions differ from Models, or how to structure a system monitoring 50 sensors.
It gives you a synthesized answer with references to the specific pages that
support it—no generic IoT knowledge, no guesswork.
Hub (Coming soon)
Natural language to deployed LoT logic, directly in your browser.
AI assistance is coming to the Coreflux Hub. Describe what you want your
broker to do, and the Hub will generate the LoT code and deploy it
all in a single browser window.
No editor. No config. Just describe and deploy.
View all posts

Key Differentiators

Grounded in the Real Docs

The Coreflux MCP Server exposes the full official documentation as a set of
searchable tools. When your AI assistant answers a Coreflux question, it reads
the documentation first. You get source references you can verify, not confident
hallucinations you can't.

Works With the AI Tools You Already Use

No new platform to learn. No proprietary chat interface to switch to.
The Coreflux MCP works with Claude, GitHub Copilot, Cursor, and any other
MCP-compatible client. One configuration entry and your existing assistant
becomes a Coreflux expert.

LoT Is Designed to Be AI-Readable

LoT's near-English syntax makes it one of the most AI-friendly automation
languages available. AI models can read it, generate it, and review it
with high accuracy, especially when combined with the MCP documentation
connection. The language and the AI layer were built to work together.

AGENTS.md: Project-Level AI Alignment

Drop an AGENTS.md file in your project root and every AI assistant that works
with your codebase learns your naming conventions, topic hierarchy, code standards, and team rules instantly. One file. Consistent AI-generated code across your
entire project.

Ask the Docs Assistant Anything

The Docs AI Assistant at docs.coreflux.org answers questions about Coreflux
in natural language. It knows the difference between Actions and Models,
can compare Route types, and will walk you through configuration step by step, citing the documentation pages it draws from.
View all posts
Try it Now

FAQ

Frequently asked questions.
What is Coreflux?
Coreflux is an Industrial IoT platform built around a high-performance MQTT broker with built-in edge computing capabilities. Unlike traditional brokers that only route messages, Coreflux processes data where it is generated using LoT (Language of Things), a human-readable automation language. The platform connects to industrial protocols like Modbus, OPC UA, Siemens S7, and Allen-Bradley, while routing data to databases, REST APIs, cloud services, and other MQTT brokers. Coreflux runs on Windows, Linux, Docker, Raspberry Pi, and cloud environments, making it suitable for edge deployments, factory floors, and enterprise-scale IoT architectures.
What is an MQTT broker and why do I need one?
An MQTT broker is the central hub that enables devices and applications to communicate using the publish-subscribe messaging pattern. When a sensor publishes data to a topic, the broker delivers that message to every application subscribed to it. This decouples your devices from your applications, meaning you can add new sensors, dashboards, or analytics tools without rewiring your entire system. For industrial IoT, an MQTT broker is the foundation of a Unified Namespace (UNS), allowing real-time data flow between PLCs, sensors, databases, and cloud platforms.
What is the Language of Things (LoT)?
LoT is a human-readable language for IoT automation created by Coreflux. It uses near-English syntax to define logic, data structures, and integrations directly within the MQTT broker. Think of LoT as SQL for real-time MQTT data. Just as SQL transforms database records, LoT transforms live MQTT streams without deploying external services. With LoT, you can define Actions (event-driven logic triggered by time or topics), Models (structured JSON schemas), Routes (connections to databases, APIs, and industrial equipment), and Rules (access control for topics and operations).
What industrial protocols does Coreflux support?
Coreflux provides native connectivity to major industrial protocols through its Routes system. Supported protocols include Modbus TCP/RTU for sensors and legacy equipment, OPC UA for modern industrial systems, Siemens S7 for S7-300, S7-400, S7-1200, and S7-1500 PLCs, Allen-Bradley for ControlLogix and CompactLogix controllers, and FINS for Omron PLCs. These protocol connectors run inside the broker, eliminating the need for separate gateway software. Data from PLCs flows directly into your MQTT namespace and can be processed, transformed, or routed to databases in real time.
How is Coreflux different from other MQTT brokers?
Traditional MQTT brokers like Mosquitto act as passive message routers. Coreflux is an active data processing platform. It combines the MQTT broker with edge computing capabilities, so you can transform, filter, validate, and route data without external services. Key differentiators include built-in LoT runtime for automation logic, native industrial protocol support without additional gateways, direct database routes to PostgreSQL, MongoDB, InfluxDB, CrateDB, and OpenSearch, Python integration for complex calculations and external libraries, and a lightweight footprint suitable for edge devices and Raspberry Pi. This means fewer moving parts, lower latency, and simpler architecture compared to assembling separate tools for connectivity, transformation, and storage.
Can I extend Coreflux with Python?
Yes. While LoT handles most IoT automation elegantly, some tasks need the full power of a programming language. Coreflux allows you to call Python functions directly from your LoT actions. Your Python code runs inside the broker alongside your automation logic, with no external services, network calls, or containers required. Use Python for complex calculations like statistics and machine learning, regex and advanced string parsing, data validation with complex rules, and external library calls using numpy, pandas, or custom packages.