@issu

In this session I was working on my project Makefast AI, which is a small AI‑powered CLI tool for developers. It runs in the terminal and helps with early project stuff like generating ideas, trying names, writing landing page copy, suggesting tech stacks, scaffolding projects, brainstorming, pitch outlines, tweets, and README drafts. I wanted something that is useful before the “real build” starts.

What I did in this session
During this time I mainly focused on making the CLI commands work properly and feel usable:

I set up and tested commands like makefast idea, makefast copy, makefast ui, makefast stack, makefast scaffold, makefast pitch, makefast tweet, makefast name, makefast validate, makefast readme, makefast history, and makefast config. Each command is small on purpose, I didn’t want one big “do everything” command.

I connected the CLI to the Groq AI API, so it can send prompts from Node.js and show the AI responses directly in the terminal for ideas, copy, stack suggestions, scaffolding, etc. This was new for me so I had to try and fix a few things.

I added local storage using a ~/.makefast folder where history, config, and logs are saved as JSON files, so the tool remembers past runs and feels more persistent.

I tested different ways to run the tool: directly with node index.js, using npm scripts (npm run dev, npm start), and using npm link so I can type makefast idea like a normal command. I also prepared it for global install (npm install -g makefast-ai).

I updated the README to explain setup (Node, npm, Groq API key, .env file), how to run commands, what scaffold templates exist (like nextjs, SaaS, api, mobile, cli, landing), and some common issues and fixes.

Problems I ran into……..
I hit a few confusing issues:

API key / .env problems:
When the Groq API key was missing or .env was not loaded, most commands failed immediately. I had to figure out why process.env.GROQ_API_KEY was undefined and learn that the way you load env variables is different in CommonJS vs ESM.

CLI command not working globally:
Sometimes after installing, typing makefast did nothing. I learned that npm link, checking the bin field in package.json, and restarting the terminal can fix path issues for global CLI commands.

Name confusion:
When the GitHub repo name and npm package/CLI name were not consistent, it was very confusing. I had to clean up naming and mention this in the README.

JSON errors:
If the history/config JSON file was missing or broken, simple JSON.parse could crash the app. I had to add checks and default values so the CLI doesn’t explode when there is bad data.

What I learned
Before this project I only knew basic Node scripts. With Makefast AI and this session, I learned:
1.how to create CLI commands and run them like makefast idea in the terminal
2.how to connect a Node CLI to an AI API and send prompts from the terminal
3.how to use .env to keep my API key outside of the code and load it safely
4.how to read and write JSON files to store history and config locally
5.how to split the project into smaller files instead of one big file
6.how to test a CLI using node index.js, npm scripts, and npm link

how to design prompts for different use cases ideas, copy, stack, scaffold, pitch, tweets so the AI output is actually useful for builders
The code might not be perfect because I am still learning, but I am happy that I built my first real CLI tool that mixes AI with developer workflow and that I can actually use myself when starting new ideas.

added cozy, cyberpunk and cartoon themes. replaced the pointer with the mob. updated the overall css by chnaging the cleaning area, added borders, modified screen styles and shades. also updated the js with the character movemnet refinement, increment in delay time, move mevemnet smoothnes.

This is a small fun web app I made using only HTML, CSS and JavaScript. In this app there is a tiny cleaner character who tries to clean a dirty foggy screen with dust, drops and fingerprints, and when the user moves the mouse the screen gets wiped.

What is inside the app

foggy screen effect with gradients, blur and low‑contrast colours dust and dirt spots sitting on top of the fog water drops and fingerprints drawn on the canva swipe effect with mouse so the dirty layer gets erased as you move funny cleaner character who walks, cleans, dances, slips, sleeps and gets angryrandom funny events with small speech bubbles and jokes3 themes: cozy, cyberpunk and cartoon, changed by CSS variables and JS buttons to clean, add dirt, change theme and trigger events responsive layout that works on desktop and mobile screens.

What I learned from this project

From this project I learned how to connect HTML, CSS and JavaScript together for one web app, instead of writing them separately.
I learned how to use more advanced gradients, shadows and colours in CSS to create a foggy dirty screen feeling. I also learned how to add images with the img tag in HTML and then reuse those same image files in JavaScript for animations.

I practiced changing the background and theme using CSS variables and JS, so I can switch between cozy, cyberpunk and cartoon styles with simple code. Another big thing I learned was how to use the canvas: drawing fog, dust and water drops there and then erasing them with the mouse to make a wipe effect. I moved the cleaner character on the screen and changed its images for different moods like walking, cleaning, dancing, slipping, sleeping, drinking coffee and being angry.

I also showed small speech bubbles with funny text using JS and made the layout responsive so it works on both desktop and mobile. For me this project feels like a cute mini browser game, not just a normal website. It is simple to understand and fun to play, but inside it uses animation, canvas, DOM, buttons and theme changing, so beginners can study it and learn many front‑end ideas.

This is the first time in my life I tried to make such a big project in a more polished and structured way. I know it is not perfect, but I learned a lot and I tried my best to make it clean, fun and educational. I hope reviewers will enjoy playing with the foggy screen and watching the little cleaner work hard. Thank you for reading.