AI Fluency & Building · AI Integration
Go deeper. Real-world tasks that need more research, planning and a polished result.
Task 1 of 3
3 useful tools an AI can call.
What to do
Build an MCP server with 3 tools for a small business (for example: look up a product, check stock in a CSV, create a to-do). Publish it on GitHub with setup instructions.
You’ll make
Done looks like
An MCP server on GitHub with 3 tools and setup instructions.
Task 2 of 3
Turn a blog post into social posts automatically.
What to do
Build an n8n or Make automation that takes each new blog post from an RSS feed, sends it to an AI to write 3 social captions, and posts them to a Slack, Discord or Telegram channel for approval.
You’ll make
Done looks like
A working automation that delivers AI-written captions for each new post.
Task 3 of 3
Wrap a data file in an API and use function calling.
What to do
Wrap a business data file (a JSON or CSV menu, product list or price list) in a small FastAPI service, and connect an AI to it with function calling so it can answer customer questions.
You’ll make
Done looks like
A small API and a demo of the AI answering customer questions from it.
Learn before you build
Free course
Free courses, including AI Fluency and Model Context Protocol, with certificates.
Free course
Learn to explore, test and document APIs.
Free tool
Open-source workflow automation you can self-host for free.
Code repo
Real MCP servers to read, run and learn from.
Article guide
The official MCP spec, quickstarts and SDKs.
Article guide
The fastest way to build a clean Python API.