MCP attempts to standardize how LLM tools like Claude Desktop and Windsurf connect with external data sources and tools. So, naturally, I thought, “Put a bird on it.” Er, I mean, “Let’s CheerLight it.” By building a CheerLights interface, I learned everything from how to build Python libraries to Chrome extensions to Adrino integrations.

This looks like a big deal. I recommend that you learn how to build your own MCP server and/or use one of the hundreds that have already been created: https://github.com/punkpeye/awesome-mcp-servers

I am using Windsurf on Windows for this tutorial.

From Windsurf, open the Terminal (View -> Terminal). Create a new folder called, “cheerlights-mcp” and enter it. Create a new file called, “server.py” and copy in the CheerLights MCP server code from GitHub.

From the terminal:

• PS C:\github> cd .\cheerlights-mcp\
• PS C:\github\cheerlights-mcp> python -m venv .venv
• PS C:\github\cheerlights-mcp> .venv\Scripts\activate
• PS C:\github\cheerlights-mcp> pip install mcp httpx

Add the CheerLights MCP to the Windsurf config file (adjust the path as needed).

{
    "mcpServers": {
        "cheerlights": {
            "command": "python",
            "args": ["C:\\github\\cheerlights-mcp\\server.py"]
        }
    }
}

Restart Windsurf.

From the terminal, start the Python server for the MCP: PS C:\github\cheerlights-mcp> python server.py

This should enable the CheerLights MCP and show two tools in the Cascade Chat area.

To use CheerLights MCP, you can ask Cascade, “What’s the current CheerLights color?” This will eventually get sent to the CheerLights MCP and the get_current_color() tool. If it all works, you will find out the latest color is magenta (or whatever the latest color happens to be)!

All of the source code is available on GitHub. Let me know if you give this a try and if you build your own. Have fun!

Arduino Library Installation

1. Open the Arduino IDE.
2. Select Tools > Manage Libraries.
3. Search for “CheerLights” and click “install”.

CheerLights Arduino Methods

• begin(const char* ssid, const char* password): Initialize the library with your WiFi credentials to let the library handle the WiFi connection.
• begin(): Initialize the library without WiFi credentials (use this to manually connect to the WiFi network and manage the WiFi connection outside the library).
• reconnect(): Reconnect to the WiFi network.
• isConnected(): Returns a boolean indicating whether the library is connected to the WiFi network.
• getCurrentColor(): Get the current CheerLights color from ThingSpeak channel 1417. Returns a pointer to a constant char array.
• currentColorName(): The current CheerLights color name (e.g., “red”). Returns a pointer to a constant char array.
• currentColorHex(): The current CheerLights color as a hex value (e.g., 0xFF0000). Returns a uint32_t.
• currentRed(), currentGreen(), currentBlue(): The RGB values for the current CheerLights color (e.g., 255, 0, 0). Returns a uint8_t.
• currentHue(), currentSaturation(), currentValue(): The HSV values for the current CheerLights color (e.g. 0, 255, 255). Returns a uint16_t for hue and uint8_t for saturation and value.
• hasColorChanged(): Returns a boolean indicating whether the current CheerLights color has changed since the last call to this method.

CheerLights Projects

Now, that you know about the Arduino library for CheerLights, you might be thinking about what you can do with it. CheerLights is a global network of synchronized lights. The lights are all set to the same color. Anyone can change the color by changing it on the CheerLights Discord Server and other social networks like Mastodon and BlueSky.

Get inspired by taking a look at the build gallery at the CheerLights website.

WLED Installation on Windows

• Download ‘esptool.exe’ app from the esptool releases area.
• Download the latest release of WLED for the device that you are using.
• Plug in the device and note the COM port.
• Open the Command Prompt and enter this command:

esptool.exe -p COMXXXXX write_flash 0x010000 WLED_0.14.4_ESP32-C3.bin

Note: Change COMXXXXX to the actual COM port that your device is connected to. If your ESP device is on COM 9, then this would be COM9.

Here’s what it looks like when everything works successfully on Windows:

WLED Configuration

This device’s button is on GPIO 9, and the data pin for LED lights is on GPIO 10. I need to set these in the WLED configuration for the lights to work.

• From WLED, select Config -> LED Preferences.
• Under LED outputs, enter ’10’ into the GPIO box.

• Scroll down the GPIO button sections.
• Select 9 for Button 0 and set to pushbutton.

• Scroll to the bottom and select Save.

CheerLights Settings (optional)

If (and when) you want to connect this WLED controller to CheerLights, just add the CheerLights MQTT server address to the Sync Interfaces. For MQTT to work, you will need to cycle the power going to the board (i.e., unplug and plug back in).

We are experiencing the capabilities and limitations of AI collectively. After OpenAI’s Dev Day, I have been thinking of ways to learn the new features, services, and APIs that were dropped. This time of year always brings about a lot of inspiring CheerLights projects and new people joining CheerLights. This also comes with a lot of questions, as the new folks are learning IoT, looking for sample code, or just looking for inspiration on what to build. This gave me an idea and a challenge. Could I build an AI chatbot that can help onboard new people onto the CheerLights project and help them learn IoT?

If you have ChatGPT Plus, you can try the CheerLights IoT Expert bot right now.

What is CheerLights?

CheerLights is an Internet of Things (IoT) project that creates a global network of lights that change color in unison. The concept is straightforward yet powerful: any connected light anywhere in the world will change to the same color when a new color is set. This allows people across different locations to feel connected through a shared lighting experience.

Features of the CheerLights IoT Expert GPT

The CheerLights IoT Expert GPT has a lot of general knowledge of Arduino, ESP Wi-Fi devices, ThingSpeak, and CheerLights. You can ask anything from the basics to advanced things like code generation for a specific device configuration.

• Natural language interface
• Onboarding prompts to get started
• Technical support and debugging
• Idea generation
• Code generation
• The world’s most sophisticated CheerLights color checker

How it works

A Generative Pre-trained Transformer (GPT) is an advanced type of artificial intelligence model designed to understand and generate human-like text. This is the technology that ChatGPT is built on. It’s built using a neural network architecture called a transformer, which excels at processing sequential data like language. GPT models are trained on vast amounts of text data, enabling them to learn language patterns, grammar, and context. They function by predicting the next word in a sequence, considering the words that have already been presented. This capability allows them to generate coherent and contextually relevant text responses. The more they are trained with diverse datasets, the better they become at mimicking human-like writing, answering questions, and even engaging in conversations, making them powerful tools in natural language processing.

OpenAI released a new feature where you can take the power of a foundational language model and direct it towards a specific domain. I created a GPT using all of the CheerLights documentation and blog posts that I have written over the past 12 years. I gave it additional instructions about where to find the latest CheerLights color and the fact that we no longer use Twitter/X. GPT-4 still thinks that you can set the color of everyone’s lights using Twitter (insert sad face emoji), but we are now on a Discord Server!

How to use the CheerLights IoT Expert Chatbot

If you are a ChatGPT Plus subscriber, all you have to do is load up the GPT by clicking this URL: https://chat.openai.com/g/g-E2zjJXE56-cheerlights-iot-expert

From there, just click on an onboarding prompt or type in a message.

Use Cases

Try one of these prompts to help inspire how to use the CheerLights IoT Expert chatbot.

• How do I tell which pin is the power pin on a WS2812B LED strip?
• I want a sparkle animation for a NeoPixel strip of 16 LEDs.
• How do I teach IoT to a group of kids using CheerLights?
• How do I show the latest color of CheerLights in my Chrome browser?
• Write Python code that retrieves the last 100 color changes and analyzes the data for any patterns.
• What are five ideas for a CheerLights display?

Get Started

Give the CheerLights IoT Expert bot a try. It can get things wrong, but it can also surprise you with something that you never thought of before. If it answers something incorrectly, let me know. I can always make changes and provide up-to-date information to make sure the bot gives the best answer possible. I am looking forward to seeing the creative CheerLights displays this season enhanced by ChatGPT. Join the CheerLights Discord server and share all of your awesome projects, ideas, and experiences.

Bonus Prompt

I enabled image creation with DALL-E 3, just for fun. You can ask the CheerLights IoT Expert GPT to create images inspired by the latest CheerLights color. Try this prompt: Create an illustration based on the latest CheerLights color, depicting a globe with electric lights.

Twitter/X

According to my own History of the CheerLights Project, I created this project around the idea that anyone on Twitter, ahem, X, could change everyone’s lights. Twitter made the most sense. It was open and had a global reach. And the bonus was that followers of people who joined the project would see the tweets, which caused some of them to join the project. It was a nice, virtuous circle that fueled the growth of CheerLights year over year. 2022 was the biggest year for the project, just hitting its peak at the same time Twitter started to crack. The API access that we have had for over 12 years was cut off, users moved to decentralized platforms, and X’s relevance has been in constant question. Don’t get me wrong. I will always have a special place in my heart for Twitter. I can’t unwind Twitter’s on my professional growth over the last 15 years.

Discord

The CheerLights community got together last year and created a Discord server bot and even released the bot’s source code on GitHub. We did this to increase the number of ways to control the CheerLights network, but as it turned out, this is now the primary way to control CheerLights. Hundreds of our users moved over to the Discord server from Twitter/X. In some ways, Discord is way better. I love being able to chat in real-time with other people who are building CheerLights. A lot of people use Discord. And Discord is constantly evolving and keeping up with the trends by adding admin and community features. In other ways, it is not ideal for CheerLights. Discord servers are not easily discoverable. You have to be in the know. It is hard to attract new users since new users don’t see public mentions of the project. But I am not aiming for perfection. I am aiming for a path forward to keep the project growing. Discord is where we are now. Join us.

How can you help?

Since our Discord server is hidden from public view, we could use your help by making CheerLights discoverable. If you share your project on your blog or social channels, include a link to the Discord server invitation. H

CheerLights Discord Server Invitation Link: https://cheerlights.com/discord

Thank you to everyone who has helped build and grow this project over the last 12 years. This is always my favorite time of year getting to see the creative ways CheerLights gets used and how it enables learning of core IoT concepts.

CheerLights is a fun way to get started with IoT, microcontrollers, and LED lights. If you know how to make CheerLights work, you can build some other awesome IoT projects. CheerLights is a great project to get people interested in IoT. It’s often used as an “Hello, World” project for the Internet of Things, and what you learn can be applied to other project ideas.

To make a Halloween-inspired project, you can take the CheerLights color and illuminate a pumpkin or even a… skull from Pimoroni.

Check out other Halloween CheerLights projects over at the CheerLights blog. Have fun with the idea, build something spooky, and help spread the… fear.

CheerLights

CheerLights is an “Internet of Things” project I created in 2011 to help teach people IoT concepts. I consider it a “Hello, World” project to learn IoT protocols, microcontroller programming, and LED light control. CheerLights is a global network of synchronized colored lights. Everyone subscribes to the latest CheerLights color, which sets their lights to the same color. The color is set on the CheerLights Discord Server by anyone.

Apps

SkyStream: Track Bluesky Hashtags

IoT Debugger: Explore IoT Data

Latest IoT Posts

• 
  Learn How to Create Your Own MCP Server for Claude Desktop and Windsurf
• 
  Official Arduino Library for CheerLights
• 
  Updating WLED Binary on ESP32-C3 Devices
• 
  Learning How to Use the Arduino UNO R4 WiFi Board
• 
  World’s First GPT for IoT? Learn IoT From the CheerLights Expert GPT powered by ChatGPT.
• 
  Moving away from X/Twitter: CheerLights is now a Discord IoT project
• 
  CheerLights for Halloween
• 
  ChatGPT Code Interpreter for ThingSpeak IoT Data Analysis
• 
  Top 10 IoT Project Ideas For ThingSpeak
• 
  Happy Valentine’s Day! Here’s some IoT.
• 
  On Air Light 2023: Easy DIY Status Light for Microsoft Teams and Zoom
• 
  Send IoT Data to ThingSpeak using AT Commands (Arduino and SIM800L Tutorial)
• 
  Give the Gift of CheerLights this Holiday Season
• 
  Internet-connected Lofi Girl Shadowbox
• 
  #FearLights: CheerLights for Halloween
• 
  The IoT Debugger App for ThingSpeak Now Includes a CheerLights Example
• 
  You Can See the Earth Move Thanks to a Seismograph Board for the Raspberry Pi
• 
  Early Experimentation: Interactive Twitch Live Streams, AI Produced Lo-Fi Music, and IoT
• 
  Updates to the CheerLights IoT Project: Node-RED, JavaScript, Twitch, and Discord
• 
  New CheerLights Installation at MathWorks Lakeside Campus
• 
  Worldwide Network of Lights: CheerLights Turns 10
• 
  The Internet of Pinball: What an IoT Expert Has to Say About Stern Pinball’s Insider Connected Platform
• 
  No Code: Interactive, IoT-enabled Halloween Pumpkin RGB Jack-o’-Lantern
• 
  No Code: On Air Light Controller For Live Streaming and Online Meetings
• 
  My Toaster is Now on TikTok
• 
  The Data Science (and Some IoT) Behind Great BBQ
• 
  Predictive Maintenance: Going Next Level With the Ninja Foodi Air Fryer
• 
  The Internet of Nouns: How Twitter and Foursquare Changed My Life
• 
  CheerLights for Raspberry Pi Zero W and Blinkt!
• 
  Adafruit Matrix Portal LED Display Diffused Acrylic Stand
• 
  ‘On Air’ Light for Microsoft Teams and Zoom Meetings
• 
  Real-time Colors on CheerLights
• 
  Lovebox: Internet of Things for Valentine’s Day
• 
  (My) History of the Internet of Things
• 
  Control IFTTT Webhooks with MATLAB
• 
  ThingSpeak at the Boston TechJam
• 
  Getting Ready for the Bay Area Maker Faire!
• 
  Use the New MQTT Service from ThingSpeak for Real-time CheerLights Updates
• 
  Use AWS DeepLens to Determine If Hotdog or Not Hotdog
• 
  How to Reset the Arduino MKR1000
• 
  CheerLights Robot for Your Nursery
• 
  Measure Wi-Fi Signal Levels with the ESP8266 and ThingSpeak
• Learn How to Use MQTT.fx with ThingSpeak
• Are Open Source IoT Platforms Ready for Primetime?
• 
  Twilio Rickrolls You Using Their API
• 
  MQTT is So Hot Right Now
• Open Source IoT Debug Tool for ThingSpeak and Particle
• 
  ThingSpeak + Particle Photon using Webhooks
• 
  Particle Electron and Webhooks to ThingSpeak!
• Particle Electron Kickstarter Kit Just Arrived
• Seeeduino Cloud Starter Kit for AWS IoT
• ESP-210 Windows USB Driver
• 
  CheerLights – What Will You Build?
• 
  CheerLights with LIFX Wi-Fi Light Bulbs
• The Washington Post IoT Infographic
• 
  Send Your Windows Server’s Disk Free Space to ThingSpeak Using PowerShell
• 
  CheerLights is Back!
• IoT at the Small Batch Beer Brewery in Winston-Salem
• 
  CheerLights is Back for the Holidays
• Un grille-pain high tech qui envoie des tweets
• 
  CheerLights: my lights are linked to everyone else’s
• 
  Top 100 “Internet of Things” Thinkers
• ThingSpeak: Building My Own Twitter (for Things)
• 
  Automatic Thermostat Control Based on Location and Weather
• 
  Toaster + Twitter = Internet of Things
• 
  New Gig, Less Hotels
• 
  TouchShield Slide Two-way Communications
• 
  Internet-enabled Message Center
• 
  First External ioBridge Project: Pet Treat Dispenser

Subscribe

I am exploring AI, IoT, community, and life… and sometimes barbecue… on a weekly basis. Subscribe to keep up with my latest musings about things.

Please leave this field empty

Check your inbox or spam folder to confirm your subscription.

ThingSpeak is an Internet of Things (IoT) platform that allows you to collect, analyze, and act on data from sensors and other IoT devices. The platform provides a web-based interface for users to create and manage channels, which are virtual locations where data from sensors and other devices can be collected and analyzed using the built-in MATLAB analytics app.

With ThingSpeak, you can visualize your data in real-time, generate alerts and notifications based on pre-defined thresholds, and even automate actions based on specific conditions. The platform also includes built-in integrations with popular IoT hardware platforms, such as Arduino and Raspberry Pi, making it easy to connect your devices to the internet and start collecting data.

Top 10 IoT Project Ideas for ThingSpeak

1. Weather Station: Build a weather station using a microcontroller such as Arduino or Raspberry Pi that collects data on temperature, humidity, and air pressure, and sends it to ThingSpeak for real-time monitoring and analysis.
2. Home Automation: Use ThingSpeak to automate your home by controlling lights, fans, and other appliances based on sensor data, such as motion, temperature, or light levels.
3. Smart Agriculture: Monitor soil moisture, temperature, and other environmental variables to optimize crop growth and water usage, using ThingSpeak to collect and analyze the data.
4. Water Quality Monitoring: Monitor water quality in lakes, rivers, and other bodies of water using sensors to collect data on pH, temperature, and other parameters, and then use ThingSpeak to analyze the data and generate alerts for unsafe conditions.
5. Smart Parking: Use ThingSpeak to monitor parking spots and guide drivers to available spots using sensors and real-time data analysis.
6. Industrial Automation: Monitor machine data in a factory or other industrial setting to detect anomalies and predict equipment failures, using ThingSpeak to collect and analyze the data.
7. Healthcare Monitoring: Use ThingSpeak to monitor patients’ health data, such as heart rate, blood pressure, and oxygen levels, in real-time and alert healthcare professionals of any critical changes.
8. Air Quality Monitoring: Use ThingSpeak to collect data on air quality, including particulate matter and pollutants, and generate alerts and visualizations for high-risk areas.
9. Smart Energy: Monitor energy usage in homes or businesses and automatically adjust consumption based on sensor data, such as occupancy or sunlight levels.
10. Smart City: Use ThingSpeak to monitor and optimize traffic flow, waste management, and other city services based on sensor data from a network of IoT devices.

Subscribe to the About Things blog for more IoT news, tips, and projects. If you need help or want to share ideas, connect with me and the community on our Quandary Discord Server.

This version of the On Air Light is cheaper and easier to build than my previous versions.

The Idea

I want a light that turns on when my webcam turns on. I want the light to turn off when my camera turns off. The idea is that my camera is on when I am in an online meeting. The light is a good indicator to everyone, including myself, that I am in a meeting.

The System

In order to turn a light on and off, I need a script that monitors my webcam for activity. This script will then send the status of my webcam to ThingSpeak, a cloud IoT platform that allows you to save and retrieve sensor data. Then, the on-air light controller will read the data from ThingSpeak and turn the light on if my camera is on.

The Parts

I was able to find everything you need on Amazon to build this project.

• On Air sign
• ESP32 board
• ESP32 screw terminal board
• LED light strip

Here are some notes about these parts.

• You can use any sign that you want to light up. I chose this one since it was $15 US and looks great. It also includes a Micro USB cable that you can use for the ESP32 device.
• Screw terminals make this project a solder-free project. Some people who wrote me said that soldering was a barrier to getting started. I want this to be accessible to as many makers as possible. Screw terminals offer a quick way to break out the pins of the ESP32 board and easily connect wires.
• Only certain styles of the ESP32 board work with the screw terminals. The terminals are for the 38-pin version of the HiLetgo ESP-WROOM-32 ESP32 dev board. The two parts that I linked to on Amazon will fit together perfectly. Other ESP32 boards might not work with the screw terminal expansion board.

ThingSpeak Setup

ThingSpeak is an IoT service hosted by MathWorks. It is focused on being easy to collect data and then retrieve that data. We are going to leverage that core functionality for this project. ThingSpeak stores data in a channel and each channel has eight fields.

Visit ThingSpeak.com and Sign Up for an account. This will just take a minute and user accounts are free. Once you have a user account, you need to create a channel. ThingSpeak channels are where data gets stored. Create a new channel by selecting Channels, My Channels, and then New Channel. Name the channel, “On Air Light” and name Field 1, “Webcam State”. Click “Save Channel” at the bottom to finish the process.

Click API Keys and note your write API key and read API key. The write API key will be used by the code to track your webcam state. The read API key will be used to turn on or off the light based on the webcam state.

Track Webcam State

I wrote a Python script that uses OpenCV to detect if the camera is in use or not. It sends a “1” to field1 of the ThingSpeak Channel if the camera is in use. It sends a 0 if the camera is not in use. Just run this code on a periodic basis to track the state of your webcam.

The code to track my webcam state and upload the state to ThingSpeak is available on GitHub. Update the code with your ThingSpeak Channel’s write API key.

Most of the magic is here:

def returnWebcamStatus(webcamIndex):

    webcam = cv2.VideoCapture(webcamIndex, cv2.CAP_DSHOW)

    if webcam.isOpened():
        webcam.release()
        return True #Webcam not in use
    else:
        return False #Webcam in use

Assembly

Get your parts together and get ready to put everything together. Take your time and step through. There is no soldering, just some connections to be made on a set of screw terminals. You will need cutters, a small screwdriver, and some tape to complete the assembly.

Connect the ESP32 board to the SCrew Terminal Board

Make sure to get the right orientation of the boards together. The 5V pins should line up and the USB part should be pointing down.

Cut Off the Out Connector from the LED Strip

We can use the connector on the out side of the LED strip. Cut this off and strip back some of the insulation. We will use this connector for the screw terminals.

Connect the Connector to the Screw Terminal Board

Connect the red wire to the 5v terminal. Connect the green wire to P13. Connect the white wire to GND.

Connect the LED Strip to the Connector

Program the ESP32

Using the Micro USB cable that comes with the On Air Light, connect the ESP32 to your computer.

The code that the Arduino IDE uses is called a “sketch” – this is just a short program that the device runs over and over. In this project, we are going to have the code that checks the ThingSpeak channel for the webcam status and sets the LED strip to either on or off. Copy the example code to your Arduino IDE and change some of the defaults to match your Wi-Fi network and ThingSpeak settings.

Once everything is set, click Sketch and then Upload. This will take the code and program the ESP32 with it. It takes a minute, so be patient. If anything goes wrong, make sure that you have the right board settings and that your “Port” matches what your laptop thinks the port is.

Open the On Air Light

The housing of the On Air light is held together by several tabs. Take your screwdriver and gently twist in each opening. It will come apart without damage. Just take your time and open each tab.

Remove the Existing LED Board

Remove the existing board on the inside of the light. Trim the wires off close to the circuit board so you have some length to work with.

Connect the USB Wires to the Board

Connect the red wire to the 5v screw terminal. Connect the black wire to the GND screw terminal. You will have two wires in each. I used some tape to make sure that the screw terminal board doesn’t move around.

Snap It All Back Together

Carefully snap the two halves back together.

Profit.

Enjoy your On Air light. Maybe this will remind you or someone in your family that you actually work when you are home.

My goal was to show you the steps as clearly as possible and share the code. I hope that you can reproduce this as easily as possible. If you need help leave a comment or join me on my new Discord server.