Four Typos Was His System: How He Built the Prompt Tool 2,000 People Actually Use
From "Replyyyy" to Right-Click: How He Turned a Ridiculous Workaround Into 2,000 Users
Published: January 23, 2026 URL: https://buildtolaunch.ai/p/prompt-management-chrome-extension-right-click-build-to-launch-friday Engagement: 38 likes, 16 comments, 4 restacks Word count: 4664
Welcome to Build to Launch Fridays, where we meet the builders turning domain expertise into AI-powered products.
Every Friday, I’m spotlighting someone from the vibe coding builders collection who’s doing exactly what I believe is the future: using AI not as just another tool, but as a true collaborator to transform curiosity, passion, and years of professional knowledge into something scalable and ownable. No VC funding, no technical co-founders, no permission required, just domain experts who decided to build.
Today, meet Kamil — the man whose prompt “system” was 20+ typos he had to memorize.
Where do you store your prompts?
For anyone who uses AI daily, you know the chaos. Notion pages you’ll never open again. Google Docs buried in folders. Notes you made once and forgot. Kamil had his own ridiculous workaround. You’ll hear what it was when he tells you himself.
Here’s what makes this interesting: Kamil consults for enterprise companies on AI adoption. He’s the guy who teaches others how to work smarter with these tools. And he couldn’t solve his own prompt problem.
So he built Right Click Prompt — a Chrome extension that lives in your browser’s context menu. Select a text field, right-click, pick a prompt. Now over 2,000 people use it.
When I tested it, I was surprised by how polished everything felt. The onboarding sends you straight to a video tutorial. Demo prompts are waiting when you log in. There’s analytics tracking how much time you’re saving. For something built by a non-coder, it didn’t feel like one.
I wanted to know how that happened.
You can explore Right Click Prompt on Kamil’s Vibe Coding Builders profile.
Quick Stats
Prompt library: 26 folders, 151 prompts (uses ~20 daily) Users: 2,000+ on Chrome extension Time to V1: ~3 weeks to Chrome Web Store Time to V2: ~1 month (mid-Dec 2025 to mid-Jan 2026) Coding background: Last coded HTML in 2001 AI stack: Claude Code (breakthrough), previously Cursor + Lovable
👉 Visit Right Click Prompt | View on Vibe Coding Builders
Origin & The Problem
What inspired you to build this product?
I help Fortune 500 companies turn AI panic into culture. My job is teaching people to think with AI, not just automate tasks away.
But here’s the embarrassing part: I couldn’t keep track of my own prompts.
I had “replyy”, “replyyy”, “replyyyy”, and “replyyyyy” as Mac text replacements, each expanding into a different reply style. That was my “system.” Four typos I had to memorize.
I tried Notion. I tried docs. But every time I needed a prompt mid-workflow, I’d have to leave what I was doing, hunt it down, and break my flow. For someone who teaches AI productivity, this was ridiculous.
The breaking point? When the time it took to remember and look up my saved prompts started exceeding the time it would take to just write a new one from scratch. At that point, what’s the point of saving them?
The insight was simple: prompts should live where you work, one right-click away, directly in the browser. I couldn’t find anything that did this well, so I built it.
*Later on, Later on, he told me he accumulated 20+ of those ridiculous memorizations — “tiktokkk”, “replyyyy”, and so on. Here’s one he kindly shared: *
**tiktokkk** →
Turn this topic into a 140-word script, starting with three strong strung together hooks (examples in knowledgebase) that capture attention and go straight to the heart of the topic with a bold, relatable statement. Follow that section with "My Name is Kamil Banc and this is AI in 60 seconds." Introduce a relevant challenge or pain point that viewers can identify with. Highlight the real consequences of not addressing this issue, adding urgency. Present a clear solution, with practical, actionable steps or a quick breakdown that viewers can apply immediately to see results or understand AI's value. Use active voice and simple, straightforward language, with short, punchy sentences. Eliminate unnecessary words and simplify complex terms to keep the script dynamic and relatable. Focus on delivering practical insights that leave the viewer with something they can use right away.
*Don't add a summary at the end. Keep the original meaning but express it in a more conversational, straightforward way.*
I’m grateful he’s open about this. It’s hard to admit when you’ve got a problem in the exact domain you’re supposed to be an expert in — like a productivity consultant who can’t find their own notes. But honestly? I’ve had equally ridiculous systems. We all do. The difference is Kamil actually fixed his.
When did the right-click interaction model click for you?
The core insight was that prompts should live where you work, right in the browser window. A prompt library that’s directly accessible in the right-click context menu was the perfect solution because it doesn’t interrupt your flow. You’re already in the browser, already in the text field, and one right-click later you’re done.
I couldn’t find anything that took this approach, so I built it.
When I tested it, that’s exactly what struck me. So clever and so clean. Most prompt tools feel like extra work — another tab to open, another app to context-switch into. This just lives where you already are.
The YouTube tutorial redirect on sign-in: is that something you developed intentionally?
It just clears up a lot of questions right away. I’m not a UI/UX designer, and maybe some of the things I do aren’t as intuitive as I think. It’s always better to show, not tell, especially showing how I’m actually interacting with the UI.
I’ll probably introduce more video throughout the product. When you can’t be there to explain something, a 2-minute video does the work for you.
That’s totally fair. Most builders (me included) don’t give visual walkthroughs for our products — we assume people will figure it out. Kamil assumes they won’t, and builds the guidance right in. Good wake-up call.
How did you decide which demo prompts to include?
They’re prompts I actually use. The “From the Creator” pack includes templates for different workflows: replies, summaries, analysis, brainstorming. The goal was to give everyone a starting point so they immediately see the value, rather than staring at an empty dashboard wondering what to do next.
Yet another product strategy I’m picking up from this conversation. Don’t make people start from zero.
Building with AI (The Vibe Coding Journey)
What was your skill level before building Right Click Prompt?
Zero. The last time I coded anything was an HTML website in a text editor in 2001. I’m a lifelong geek (I was a photographer, a 3D artist) but never had any formal technical training. Absolutely no coding experience before RCP.
2001. That’s 24 years without touching code. And now he’s shipping polished software. It really does make you wonder what else becomes possible when the barriers drop like this.
Walk me through the story. What was the very first MVP?
The first version could basically just store prompts locally with folders. That’s it. No cloud sync, no sharing, no dashboard. Just a Chrome extension that saved prompts to your browser’s local storage.
I was so excited when I got drag-and-drop reordering to work. I thought it was the coolest thing ever. It blew my mind that I could build something interactive like that.
The joy in this answer. Drag-and-drop might seem naive to someone more experienced, but that feeling when something actually responds to your input and works? Still pure magic.
What was the hardest part of building it?
I hadn’t written code since building an HTML website in a text editor in 2001.
The first version in early 2025? I didn’t even know to save my progress to GitHub. I made changes to something that worked... and it collapsed. Had to rebuild everything from scratch.
But the real wall came in summer 2025 when I wanted to add cloud sync and team sharing. I tried Cursor, tried Lovable, but I genuinely didn’t understand how a dashboard would even talk to a Chrome extension. The questions I needed to ask were so basic I didn’t know how to ask them.
That’s when I almost quit. I have a consulting business. I’m not exactly sitting around with free time.
Then someone convinced me to try Claude Code in the terminal. Everything clicked. What took months of frustration before now happens in focused sessions. From mid-December to mid-January (about one month) I went from nothing to a polished beta. The changelog tells the story: rightclickprompt.com/changelog
“The questions were so basic I didn’t know how to ask them.” I think every non-coder builder knows this wall. It’s not the technical skill that blocks you — it’s not knowing what you don’t know. The vocabulary gap is real. And that friend who talked him into trying Claude Code? Life saver. I have mixed feelings about Claude Code myself — really powerful for some work, hit-or-miss for others — but clearly it was the right fit for him.
What made you expand from that initial MVP into a full product?
I wanted cloud sync and team sharing, features that would actually help people in their workflows. V1 only ran locally in your browser memory. If you switched computers or cleared your cache, everything was gone.
I had multiple conversations with different AIs about the features I thought were important. Over time, it crystallized that the combination of a full-fledged dashboard AND an extension would be the best approach.
I see a craftsman here. Not someone just trying to ship fast, but thoughtfully working through which features actually matter for people’s day-to-day use.
What vibe coding tools did you use?
Started with Cursor when it first came out (early 2025). Also tried Lovable. But I honestly didn’t have the technical background to understand what I was building.
The breakthrough was Claude Code in the terminal. It just resonated with me. Ever since Opus 4.5 came out and I had some time in December, I was blown away by the possibilities. The incremental improvement through conversation is amazing.
I feel for him here. Early 2025 Cursor wasn’t that powerful — I had to customize a lot of tools to make it autonomous. No wonder he hit the technical barrier. If he tried Cursor now, it’d be a different story. And his timing picking up Claude Code was perfect — late 2025 is when these tools really evolved into something autonomous and adaptive. Prime time to get started.
If you want to try Claude Code yourself, I wrote a beginner’s guide to get you from zero to your first project in 30 minutes.
Different tools work for different people. He tried Cursor, tried Lovable, and it was Claude Code that finally clicked. I tell people this all the time: if one tool isn’t working, try another. What resonates for someone else might not resonate for you. (I tested 5 AI coding platforms with the same prompt — the results varied wildly.)
What was the most frustrating part? Did you almost give up?
Summer 2025. I wanted to add cloud sync and team sharing but couldn’t figure out how a dashboard would connect to an extension. The questions were so basic I didn’t know how to phrase them.
That’s when I thought maybe I should just quit and focus on my consulting business. It’s not like I’m not busy.
Luckily, someone talked me into trying Claude Code. Decided to give it one more shot, and this time it worked.
That summer 2025 moment is so real. I think every builder has a phase where they wonder if they should just focus on what’s already working instead of chasing something new. The fact that he almost quit and then found the right tool — sometimes perseverance isn’t about pushing harder, it’s about trying a different door. (If you’re bouncing between tools and feeling stuck, I wrote about breaking out of AI tool paralysis — it might help.)
Product & Design
How much of the UI came from your design background vs. AI?
I’m not a UI/UX designer. I was a photographer, did some 3D art. But I have creative judgment about how I’d like things to look.
It’s mostly my creative direction with AI doing the execution. I ask what makes logical sense, check if I’m having blind spots based on best practices. I don’t try to reinvent the wheel too much, but I tweak things until I like them.
The interface is simply so beautiful. I love how it’s rendered. Creative direction + AI execution — that’s a useful formula. You don’t need to be a designer. You need taste and the willingness to iterate until it feels right. That’s something I’m still learning myself.
How does the analytics tracking work?
Yes, built everything from scratch with AI, including a feedback system that’s basically my own version of Zendesk. Built that in about 15 minutes of talking to Claude Code.
The analytics (streak, time saved, prompts used) all run on the backend. Every prompt copy gets logged, and the dashboard calculates the stats.
The analytics are so clever — current streak, time saved, prompts used. And he built his own Zendesk in just 15 minutes. I’ll remember that next time I’m about to pay for a SaaS tool. Maybe I can just... build it?
How does right-click know where prompts are “allowed” to appear?
It works in two ways:
- Text field + right-click: If you click into a text field, right-click, and select a prompt from the context menu, it will auto-populate (if the auto-paste toggle is on). Otherwise, it copies to your clipboard.
- Anywhere else: It just copies to clipboard, and you paste wherever you want.
So it’s not that prompts are “blocked” from Google search. You just right-click anywhere, copy the prompt, and paste it. The auto-populate only works in actual text input fields.
When I tested it on ChatGPT, Claude, Perplexity — it worked beautifully, prompts organized by folders. But Google search confused me. Now I get it, auto-populate needs a text field.
Why did you choose JSON format for import/export?
Wasn’t a conscious choice initially. It’s just what made sense when building with AI. I got used to it, made it better over time, and now it works well.
I don’t waste time fixing things that work. Maybe I’ll evolve it eventually, but JSON is universal and gets the job done.
I like how confidently he makes these decisions and sticks with them. No second-guessing, no over-engineering.
I noticed you have feedback collection, bug reports, and feature requests built right into the product. What’s the ratio between bug reports and feature requests?
Looking at the changelog, feature requests outnumber bug reports, which is great. But I’m sure that ratio will shift over time as more people use it and find edge cases.
Having feedback built directly into the product is such a smart choice — you want it frictionless when someone has a thought. And feature requests outnumbering bugs? That’s the kind of problem you want to have.
Growth & Community
How long did it take to get to the first 1,000 users?
Probably around mid-2025, so about 6 months after launching V1.
I didn’t really track it closely. Growth came from a few things:
- Some people reviewing it on YouTube
- AI tool directories
- Natural discovery through the Chrome Web Store
- My personal social media channels
No ads. No strategic marketing. Just put it out there and let it grow.
I saw he’s got at least 2,000 users on the Chrome extension. That’s incredible. No ads. No strategic marketing. When you’re in the trenches experiencing the pain before everyone else, that’s your unfair advantage — you get to build before the market catches up.
What does the Discord community do?
It’s a better way for me to get direct interaction and feedback from users. Though honestly, I think a lot of my users don’t even have Discord, or at least they haven’t joined yet.
It’s not solely about Right Click Prompt. It’s also about prompt management, AI workflows, and the broader conversation around using AI effectively.
Update: Kamil is actually closing the Discord after beta, not a lot of activity beyond bug reports and feature requests. The internal feedback system he built works better anyway.
Honestly, good call. If the built-in system works better, why maintain something that doesn’t? I’ve only been in the creator world a bit over a year. Every approach he takes looks so “correct” to me.
What’s a surprising insight you had while working on the project?
I built sharing for teams, but realized creators need it even more.
Think about it: when someone shares a prompt on Substack or Twitter, what do you do? Drag your finger across the screen, copy it, then... save it where exactly? A note you’ll forget? The same problem I started with.
So I pivoted. Right-Click Prompt is becoming the best way to share prompts online, not just store them. You can create branded prompt packs, generate QR codes, and give people one-click subscribe links. I even built a custom GPT that generates importable prompt files for you.
The “aha” was realizing prompt management and prompt sharing are the same problem from different angles.
That’s a business insight I hadn’t considered. Prompt management and prompt sharing are the same problem from different angles. He’s positioning for distribution, not just storage.
Market & Competitors
Did you know about similar solutions when you started building? (e.g. Karen recently built something with the same idea)
I didn’t know about Karen’s Prompt Collector or similar tools. I honestly didn’t find what I was looking for when I searched. Nothing took the same right-click approach.
The only functionality inspiration was Grammarly. I wanted that level of interactiveness, where the tool is just there when you need it without disrupting your flow.
How do I think about competitors? There’s clearly demand in the market if multiple people are building similar things. That’s validation. I just focus on building what I need and making it better.
To be fair, Karen’s build happened very recently, so no surprises he didn’t know about it. And both approaches are pretty different. What I really like is the Grammarly inspiration — uninterrupted workflow is so important.
Did you do market research, or just build what you needed?
Just built what I needed and assumed others needed it too. I had a real problem, couldn’t find a good solution, so I made one.
The 2,000+ users suggest the assumption was right.
Absolutely. Some people do extensive research, some don’t. After seeing many domain experts build their tools, I see there’s no right way. What works for you is the right way.
Reflections & Advice
What did this build teach you about your taste as a builder?
I learned to take the Chad approach. I used to be a midwit, overthinking everything, trying to make it perfect before shipping.
Now I ship when it feels right. Before the beta launch, I obsessed over one thing: making sure nobody loses their existing prompts during migration. That mattered. Everything else could be iterated.
The line for “good enough” is: does the core functionality work well? If yes, ship it and improve based on real feedback.
“The Chad approach.” He named it. I love that. The midwit meme is so real — sometimes you just need to stop overthinking and ship. His priority was right: don’t break people’s data. Everything else is iteration. (For anyone wondering what “ready to ship” actually looks like, I have a smoke testing guide that might help.)
What question do you wish people would ask about this build?
“Can I please get verified?”
I want to start sharing my prompt packs on social media, and I’d love to have a verification badge on my profile. The verification system exists (I built it) but nobody’s asked for it yet!
Ha! You know what? I didn’t even know there’s a verification badge available. But yeah, I’m not looking for verifications either — not even on LinkedIn or X.
What’s your personal productivity stack?
- Voice dictation: Whisper, AquaVoice (talking is faster than typing)
- Claude Code: The main building tool
- Otter: Note-taking
- Calendly: Scheduling
That’s basically it. Simple stack, high output.
Glad to see some overlap here.
Do you have any mantras or productivity tips?
Voice dictation changed everything. Talking to Claude Code is faster than typing, and it keeps the momentum going.
Oh, and I built a Space Invaders game into my changelog where you fight bugs instead of aliens. Stargates are version numbers. User feedback upgrades your weapons. Someone asked me to open-source it, and I didn’t know how, but I figured it out: github.com/kbanc85/changelog-invaders
I’ve been exploring voice-first workflows too — when typing becomes the bottleneck, it’s wild how much faster you can move.
Also, you built whaat into the changelog? That’s absurd and delightful. This is the kind of whimsy that makes products memorable. With Claude Code able to do anything, injecting personality into apps — why not?
If you could go back in time, what advice would you give yourself?
Save your code to GitHub. Seriously. Day one.
This. I cannot stress this enough. It’s one of the essential software engineering practices every builder needs.
But the real advice: just ship it.
I spent months overthinking V2, bouncing between tools, trying to understand architectures I wasn’t ready for. Meanwhile, V1 was quietly getting 2,000+ users just by existing.
The beta launch was scary. I was terrified of edge cases. But the moment I launched, improvement velocity exploded. Real users give you real feedback. You can’t simulate that.
For anyone vibe-coding their first product: dive in, have fun with it, and don’t wait until it’s perfect. If you want a structured starting point, I made a prompt pack for validating ideas: rightclickprompt.com/s/497d2bed
The biggest hurdle is just starting. Everything else has been documented online forever.
V1 was gaining users while he was stuck overthinking V2. That’s the unglamorous truth about building. The product doesn’t care about your architectural debates — it just wants to exist and be useful. And get his prompt! He’s the expert of it.
What’s your #1 warning for non-coders with ideas?
Don’t get married to your idea too strongly. Keep an open mind about what you want versus what’s actually possible. Don’t get stuck forcing your vision into the final product. Sometimes we don’t know what we don’t know, especially as non-coders. Let the building process teach you.
“Let the building process teach you.” That hit me. It’s not just about the product. It’s about discovering what you’re capable of when you stop waiting for permission. This is why I keep interviewing these builders — every conversation teaches me something about myself too.
What’s your next big milestone?
Making prompt sharing actually good.
I just launched a Creator Studio where you can build prompt packs, generate importable files, and share branded links with QR codes.
Next: user profiles with all your public prompts, then a discovery engine (basically Instagram for prompts). A place to find the exact prompt template someone else already perfected.
But I’m staying disciplined. The highest-value features come first:
- Managing your prompt library seamlessly
- Keeping prompts accessible in the browser (no flow interruption)
- Sharing prompts in a way that actually works
The marketplace dreams come later.
“Instagram for prompts.” I also have the marketplace dream for vibecoding.builders. Let’s hope both our dreams come true!
Bonus: The Website Audit
I ran your website through [an audit tool I built](https://www.vibecoding.builders/custom-tools/website-auditor) for [vibe coders](https://www.vibecoding.builders/custom-tools) and found something interesting: performance is excellent (100/100), but AI discoverability is low because the site is JavaScript-heavy. AI crawlers basically see an empty page. Have you thought about SEO?
I’ll definitely work on this more. Right now the Chrome extension and dashboard go together as one product, so they need to be intertwined. It’s very important to me that the word gets out.
The JavaScript-heavy rendering is a known tradeoff with modern frameworks. I’ll address it as I go, probably adding some static content and meta tags for AI crawlers.
The irony isn’t lost on me: a prompt library tool that’s invisible to the AI systems that might benefit most from discovering it. But that’s the thing about shipping early — you learn what matters by having real users tell you. SEO wasn’t the priority when the core product needed work. Now it will be. (If you’re curious about making your product visible to AI crawlers, I wrote about SEO for AI and LLM discoverability.)
Connect & Explore
Kamil’s Platforms:
- 🚀 Right Click Prompt — Your prompts, one right-click away
- 🎨 Creator Studio — Build and share branded prompt packs
- 📋 Changelog — Every update (and the Space Invaders game)
- 👤 Kamil’s Vibe Coding Builders Profile — See his full journey
Free Resources:
- 📚 Idea Validation Prompt Pack — Prompts for validating your next idea
- 🎮 Changelog Invaders — The Space Invaders changelog game (open source)
- 📖 Development Journey — The full story
- Website audit toolkit
Previous Launch:
Vibe Coding Resources
If Kamil’s story resonated with you, here’s where to go next:
Getting Started with AI Coding:
- Claude Code for Everyone: From Zero to Your First Project — the tool that changed everything for Kamil
- I Tested 5 AI Coding Platforms With One Universal Prompt — find the right tool for you
- How AI Builders Use Cursor: The Complete Guide — if you want to explore the Cursor path
Making Your Code Production-Ready:
- Smoke Testing for Vibe Coders — the essential testing guide
- How to Make Vibe Coding Production-Ready — from “it works on my machine” to actually works
- The Essential Software Engineering Practices Every AI Builder Needs — what the pros do
Prompting & Workflows:
- Complete Guide to Prompting AI Coding Tools — every platform has a guide, this is mine
- From AI Tool Paralysis to Daily AI Workflow — 3 modes, 3 prompts
Building Smart:
- The Definitive Guide to Cost-Effective AI Building — ship without burning cash
More builder stories: Browse the full Build to Launch Friday archive
Final Thoughts
What strikes me most about Kamil isn’t the product — it’s the trajectory.
He went from “I don’t know how a dashboard talks to an extension” to shipping a polished SaaS in one month. He almost quit mid-2025, tried a different tool, and everything changed. There’s no secret framework here, no five-step process. Just someone who kept trying different doors until one opened.
I think about this with my own projects sometimes. We get so attached to our specific approach that we forget there might be a completely different path that works better for how our brains work. Kamil bounced between Cursor, Lovable, and Claude Code before finding his match. If he’d given up after the first two, Right Click Prompt wouldn’t exist.
The lesson isn’t about the tool. It’s about not letting “this isn’t working” become “I can’t do this.” They’re different problems with different solutions.
If that resonates with you, connect with Kamil. He’s the kind of builder who figures things out by doing them, and he’s genuinely generous with sharing what he’s learned.
[
AI Adopters ClubBecome the AI expert at your company. Get practical workflows and templates that save time, cut costs, make money, and prove your value. I give you the playbook, you get the results. Deal?By Kamil Banc](https://aiadopters.club?utm_source=substack&utm_campaign=publication_embed&utm_medium=web) If you’re turning your expertise into products, building with AI, or helping others do the same, you belong here. Join the vibe coding builders community and get featured on Build to Launch Friday. Curious why it all started? Here’s the full story behind Vibe Coding Builders.
Your turn:
Where are your prompts living right now? And be honest — is it some ridiculous workaround you’d be embarrassed to explain out loud? (No judgment. We’ve all been there.)
Have you ever almost quit building something, only to try one more approach that finally worked? Or are you in that almost-quitting phase right now, wondering if you should keep going?
Kamil went from a long list of memorized typos to 2,000 users. What will your builder story be?
— Jenny