A whole workflow in a shoebox (errr skillbox).
One of the best developments out of AI in the past year has been skills.
These are commands that you call on inside AI harnesses like Claude Code or Codex, and to a lesser degree Claude Cowork.
At the simplest level, a skill is a pre-written prompt. You write it once, save it as a markdown file, then call it with a slash command — /skill-name — and the AI runs the entire prompt for you. No more copy-pasting long instructions into chat windows.
But skills go further than saved prompts. You can bundle Python scripts, JavaScript, JSON schemas and other technical files alongside the markdown, turning what would be a loose AI prompt into a deterministic, repeatable process. For example, bundling a Python script that calls an image generation API so you can pass in exact parameters like resolution and output format reliably every time.
Skills can also be configured to accept what’s known as arguments — information, file references or other inputs that tell the skill to consume the input and use it while running the prompt.
You have likely already seen the skills pre-installed in your Claude Code or Codex setup. And you’ve probably seen people share them online via GitHub links.
When you see a skill on Github that you want to try, you don’t need to install it yourself — that’s the old way.
The AI-age way: paste the Github link into Claude Code and ask it to security-scan it, then install it for you.
Skills can be installed at three scopes:
- User scope makes the skill available computer-wide across every project.
- Project scope locks it to a specific folder.
- Local project scope keeps it in the project but only for you, which is useful on shared folders.
If you’ll use it across many projects, install it at user scope. If it’s specific to one project, install it at project scope. Local project scope is really only if you have a lot of people working on the same shared project.
Another mindset shift is that you should not be writing skills by hand.
Instead, ask the AI to create skills for you.
Describe what you want the skill to do, provide any constraints, and let it generate the markdown file and any bundled scripts. I recommend using the Superpowers plugin and the /writing-skills command — it walks through skill creation step by step, handles the frontmatter, and tests the result.
Bonus: My favourite skill right now is /brainstorming, also from Superpowers. It was originally designed for developers thinking through app features, but it works for EVERYTHING.
You can use it for development, for creative work, for problem solving, and even for creating new skills.
Courses
I have a range of online courses that teach business people what they need to know about productivity and AI:
1. Next Level Productivity
A practical, straightforward course that teaches you how to achieve elite-level personal productivity in today’s constantly interrupted world.
2. The AI Agents, Automations and Agentic Workflows Guide
This non-technical course shows business people and non-coders how they can build and use AI agents in ChatGPT and Zapier.
3. The Complete Claude, Claude Code & AI for Work Productivity
A comprehensive course taking you from beginner AI concepts like prompt and context engineering, to cutting-edge AI productivity using terminal-based AI tools like Claude Code for non-coding office work. Usable with ChatGPT, Claude, Gemini and other current LLMs.
4. The AI Playbook
This is my longer premium course on how businesses can deploy AI tools and technology across their processes and teams.
5. ChatGPT for Managers
See how AI can solve complex management challenges in less than 30 seconds. Full prompt library and examples included.
That’s it for this week!
— Aaron