Skip to content

List of AI Agent Skills

An AI agent skill is a folder with a SKILL.md file that teaches a coding agent, such as Claude Code, how to do one job well and the same way every time. The skills below came from this book and from the related textbooks listed in the course description. Each one turns a task that would take a club leader an afternoon into a single request: "make a challenge card for the motor lab" or "draw the wiring diagram for the Moving Rainbow kit."

To use a skill, copy or link its folder into your agent's skills directory. For Claude Code that is ~/.claude/skills/. Each entry below links to the skill's folder on GitHub, where you can read the full instructions before you install it.

Challenge Card Creator

Writes and prints the challenge cards students pick up from the center table. Each card has a front side with a title, an icon, a short description and numbered steps, and a back side with hints followed by a working solution in code or in circuit-building steps. Cards are color-coded by difficulty: green for beginner, blue for intermediate and black for advanced. The skill lays out each card as two portrait panels on one landscape 8.5 x 11 sheet, so you print on one side, fold and laminate. It also writes an image prompt for the card illustration, validates the card data, and checks that Python solutions actually run.

Scratch Lab Generator

Turns a one-sentence description of a challenge and its solution into a finished Scratch lab. Each lab includes a lesson page narrated by the mascot, a "before" and an "after" block diagram, and two real .sb3 project files students can open directly in Scratch. The block programs are written first as text, and both the diagrams and the downloadable projects are generated from that text, so they never disagree with each other.

Python Lab Generator

Generates a complete student-facing Python lab page from a topic or concept. The page has an in-browser code editor with line numbers and syntax highlighting, connected to the Skulpt Python interpreter, so students read, run and edit code without installing anything. The skill prefers turtle graphics labs that draw on screen, and falls back to text-only labs when a program produces no drawing. Each lab includes prediction prompts, learning checks and experiments, and the skill adds the new lab to the lab index page.

Project Kit Generator

Turns a dull single-sensor MicroPython lab into a complete, classroom-ready project kit. A low-cost color display (about $5) acts as a "window" on the sensor so students watch it react to their breath and fingertips. The skill builds a ladder of small numbered labs that each print TEST PASS or TEST FAIL, and it verifies every lab in a desktop simulator before any hardware exists. It also writes the student lesson, the README, and a printable box cover with a fun name, a photo and a QR code.

Circuit Diagram Generator

Creates clear wiring diagrams for microcontroller projects in the Moving Rainbow house style: a red power rail across the top, a black ground rail across the bottom, amber signal wires, light component blocks and a white background. It covers Raspberry Pi Pico projects with NeoPixels, push buttons, rotary encoders, sensors, switches and batteries. Each diagram is a small Python program built on the schemdraw library, so it is easy to tweak and re-render, and it produces both PNG and SVG files.

LED Strip Pattern Generator

Writes MicroPython programs for the Moving Rainbow LED strip project. The default hardware is a Raspberry Pi Pico, a 30-pixel NeoPixel strip and two push buttons, with pin numbers kept in a shared config.py file. The skill knows the classic patterns (moving dot, color wipe, rainbow cycle, comet tail and random effects) and can combine several of them into a multi-mode program where a button press switches between patterns.

Pi Keys Generator

Generates Python and shell scripts that control the RGB lights under the keys of the Raspberry Pi 500+ keyboard. Ask for "turn F1-F12 blue," a typing-speed indicator or an email notification flash and it writes the script using the RPiKeyboardConfig library and the rpi-keyboard-config command-line tool. It is a fun way to show students that code can change something physical on the desk in front of them.

Purchasing Guide Generator

Turns a short parts list into a teacher-ready purchasing guide for a kit. The page opens with a bill of materials that lists required parts first and optional parts after. Each part then gets its own section with a photo, what it does in the kit, any prep work such as soldering leads, a price estimate for one unit and for a class of 20, the exact search keywords that find it, and live eBay, AliExpress and Amazon search links. Use it when you need to answer "what do I order and what will it cost?" before a purchase order goes out.

Breadboard Simulation Generator

Builds interactive p5.js MicroSims drawn on a solderless breadboard. Students see what they would see on their desk: a board, components pushed into numbered holes, jumper wires, and current visibly moving along the wires when a button closes the circuit. The simulation runs a real circuit solver for LEDs, resistors, buttons, switches, capacitors, diodes, potentiometers, buzzers and NPN transistors, and can show an optional scope panel with voltage and current over time. The skill is set to run on the Opus model, which handles the layout and circuit logic more reliably.

Schemdraw Circuit Drawer

Converts a plain-language circuit description into a Python Schemdraw program and a rendered SVG or PNG schematic. It handles series, parallel and mixed circuits as well as transistor, op-amp, filter and power circuits. The skill keeps the original request in the program's docstring, runs a validation script on the output, and inspects the rendered image against the request before it reports success. Where a chapter is in scope, it proposes the best place for the diagram and asks you to confirm before editing the chapter.

Hands-On Lab Evaluator

Scores a hands-on electronics lab against a 103-point rubric built around one question: can a 13-year-old with a $50 parts kit follow the page alone, build the circuit and know why it works? The score goes into the page's frontmatter as quality_score. Every lost point becomes a specific item in the lab's TODO list, together with the name of the skill that can generate the missing piece. Run it before publishing a new lab so you know what is missing before you write anything else.

Add Robotics Sensor

Adds a new standalone sensor kit to the textbook from a sensor description and the way you actually wired it to a Raspberry Pi Pico. The skill works with the real board over mpremote: it runs I2C scans and identity checks, diagnoses real failures, and only then writes the raw-data lesson, a Thonny Plotter lesson, the upload script, the README and the published student guide. Because it starts from what works on real hardware, the guide matches what students will see. You need the board connected to your computer to use it.

Kit Quality Guide

Audits an existing robot or sensor kit, or scaffolds a new one, against a full quality checklist. It checks that each kit pairs its student lesson pages with its MicroPython code, and that the parts list and wiring photos are complete. It also checks that the config.py and upload script follow the standard conventions, that the README serves both students and instructors, and that lessons progress from simple to complex. It is the companion to Add Robotics Sensor, which handles the live hardware bring-up.

Kit Box Cover Generator

Creates a printable, single-page HTML box cover for a hardware kit. The 6 x 4 inch retail-style cover shows a real photo of the assembled kit, a short list of reasons a student would want to build it, audience badges and a footer bar of specifications. It rewrites the engineering README into student-friendly benefits instead of copying it. You must supply a photo of the assembled kit, and the skill asks for one if you have not.

MicroPython OLED Renderer

Runs MicroPython drawing code on your computer and saves what the 128 x 64 monochrome OLED screen would show as a PNG image, scaled up five times to 640 x 320 pixels. It works with unmodified framebuf and ssd1306 scripts, so you can preview a robot face or generate an image for a lesson without any hardware. For animations it can capture several frames and combine them into a GIF.

The related textbooks Clocks and Watches, Signal Processing on a $5 MicroController, STEM Classroom Administration, Raspberry Pi STEM and Learning Linux did not have any skills when this list was compiled.