Ladybug Tools MCP is ready!

share
1

Hi,community！

I developed a Ladybug Tools MCP for coding-type Agent applications, enabling the use of the core functions of Ladybug Tools through natural language. This process does not require the use of a CAD interface.

Garden
Garden1254×1254 219 KB

Several months ago, I released the beta version of Honeybee-MCP. At that time, the functionality was quite limited. Apart from editing, there didn’t seem to be much else that could be done. However, after a long period of research and testing, it has now become a complete and scalable framework, capable of utilizing the core functions of Ladybug Tools in any system.

In this project, apart from MCP, there are two other main elements, namely “Garden” and “Flowerpot”. You can learn about their actual meanings through Github.

Let’s start by giving some brief introductions！

Overview

Ladybug Tools MCP is an MCP service built with FastMCP for agent applications. Through natural-language conversation, users can use the core capabilities of Ladybug Tools for common workflows including modeling, editing, querying, simulation, and data visualization, and can do so without depending on a CAD interface.

Several months ago, I released the beta version of Honeybee-MCP. At that time, the functionality was quite limited. Apart from editing, there didn’t seem to be much else that could be done. However, after a long period of research and testing, it has now become a complete and scalable framework, capable of utilizing the core functions of Ladybug Tools in any system.

User Groups

The original purpose of Ladybug Tools MCP is to turn design or technical concepts into concrete outputs quickly. For example, when a professor explains “What is a Trombe wall?” in a building technology course, a student can open Codex voice mode during the lecture, and by the time the explanation is finished, Codex can already transform the concept into inspectable models and files, together with graphical workflow output.

How to install

I suggest that you use the intelligent agent application to perform the installation for you instead of doing it manually. These applications can be Codex or Cluade Code. If you are not a user of any intelligent agent application, I suggest you use Codex. This project is almost entirely built by Codex.

All you need to do is send this sentence in Codex. He should be able to assist you with all the tasks. You just need to have a cup of coffee and wait to use it.

Hello, Codex. Please help me install this project and configure the MCP into my agent application. https://github.com/LoftyTao/ladybug-tools-mcp

First Use

After Ladybug Tools MCP is enabled, you can invoke the ladybug-tools-mcp-use skill with /, then input HI , Ladybug Tools ! to activate the onboarding flow for the three main usage intents that we provide. After the onboarding is complete, you can start building according to your intent.

fist use

It should be noted that in this project, I have provided a usage method for the Codex Plugin. You can also activate the Ladybug Tools MCP by enabling the plugin.

@ladybug-tools-mcp Hi,Ladybug Tools!

In general, the agent application will output the onboarding template according to the guidance in our skills, but the actual result still depends on the host application’s instructions and the base capability of the language model. I strongly recommend that you use the best model available within your means in order to use our tools more effectively.

Workflow Examples

Due to space limitations, I cannot fully demonstrate to you how all the functions should be used. However, you can achieve the main functions I have described here by communicating with LLMs.

What I need to clarify is that if you want to preview everything you create for the model, please tell your agent to export it in SVG or HTML format using the Visualization Set.

Acknowledgements

Special thanks to the Ladybug Tools community and the Ladybug Tools team:

Beyond that, the implementation core of this project remains the Ladybug Tools Core SDK, which is the result of many years of development by the Ladybug Tools team.

3

Solid implementation @ZhengrongTao congrats! I have a question, can the model be prompted to use external skill from energyplus? Like create custom EMS logic or measures and use the LB built-in measure or idf injector? I assume this would need an iterative process and might be an overkill tokens-wise.

4

Hi @MohssinDr

Yes, you’re right. I postponed the plan to open additional-idf and measures because of the issue with Tokens consumption.because the custom-added fields are prone to come with large warning reports, and during the debugging process, they may cause significant consumption of tokens’ resources. Thus, I have lowered the priority of its initial release.

I have always believed that the use of MCP may still require some time for promotion and popularization. Therefore, I haven’t conducted many tests on these two options which are used by advanced users. It seems that I need to re-examine this issue. :grinning_face:

However, these two options have always been part of the main code base and have been available in my test cases. I can make the next submission within half an hour and enable these two options so that you can test these usages through this project. If you encounter any problems during the testing process, please feel free to contact me at any time.

5

@MohssinDr, for an MCP closer to the EnergyPlus schema, check this out: GitHub - idfkit/idfkit-mcp: 🤖 Model Context Protocol (MCP) server for EnergyPlus building energy simulation · GitHub (full disclosure, I’m the author). You could take your IDF model generated by @ZhengrongTao’s MCP and hand it off to idfkit for full EnergyPlus capabilities including EMS.

Other noteworthy options in this new MCP space:

PS: Super annoying, I’m limited by 2 URLs for this post… hence github-dot-com

Hope this helps!

6

Hi, all

I have made some updates recently.

Release Ladybug Tools MCP v1.0.0

for Windows only

  • Ironbug is now available
  • Fairfly is now available

Now, you can easily create a customized air conditioning system through natural language, such as primary-secondary pump variation systems, multi-loop coupled systems, etc.

For some simple building envelope systems, you should be able to easily create a heat transfer calculation model using Fairyfly through some instructions and by using MCP. Especially when you have some commonly used templates, you only need to compile them as Agent Skills, and the efficiency will be significantly higher than modeling in Grasshopper.

Furthermore, the modeling part of Dragonfly and the computing part of UWG are now ready. However, the other core computing components still require some time for development.The next step might be to gradually expand the application scope of these tools.

Enjoy it！

Zhengrong