projects

development

engineering

reflection

tools

gaming

chatbot

android

capacitor

play-store

chatuft1com

chrome-extension

chatgpt

search

browser

side-project

automation

personal-projects

devtools

productivity

chatbots

data

netlify

json

static-site

resume

openai

anthropic

other

minecraft

journey

wanabusful


**ChatGPT Search v2.0.1** is live on the [Chrome Web Store](https://chromewebstore.google.com/detail/chatgpt-search/apeckljcdhfjjanmjbceanoneldpdndm). This release repurposes the extension around what actually works in Chromium browsers today: **auto-send** when ChatGPT opens with a `?prompt=` URL, **quick search** from the popup, and a **one-time setup guide** so you can type `c` in the address bar and jump straight into ChatGPT.

**Repo:** [github.com/jovylle/add-as-search-engine](https://github.com/jovylle/add-as-search-engine)

## What changed in v2

Earlier versions leaned on omnibox APIs and redirect tricks. Browsers block extensions from adding search engines programmatically — that is a malware vector — and omnibox-only flows do not cover every Chromium browser. v2 strips that complexity and focuses on three things that ship cleanly and stay store-safe:

- **Auto-send** — open `https://chatgpt.com/?prompt=your+question` and the extension clicks Send for you (toggle in the popup)
- **Quick search** — type a prompt in the popup and go
- **Setup guide** — copy-paste values to add ChatGPT as a site search engine once (Name: `ChatGPT`, Shortcut: `c`, URL: `https://chatgpt.com/?prompt=%s`)

Removed in v2: omnibox API, `chrome_settings_overrides`, background service worker, and any owned redirect server. Queries go **directly to ChatGPT**.

## How to use it

1. Install from the [Chrome Web Store](https://chromewebstore.google.com/detail/chatgpt-search/apeckljcdhfjjanmjbceanoneldpdndm) (works in Chrome, Edge, Brave, Opera, and Vivaldi)
2. Open the extension popup and follow the **one-time setup** for your browser
3. Type `c`, press Tab, enter your question, press Enter
4. ChatGPT opens with your prompt; auto-send fires if you left it on

Or skip the shortcut and use **Quick search** in the popup whenever you want.

## Permissions (and why they are minimal)

| Permission | Purpose |
|----------|--------|
| `storage` | Saves auto-send on/off locally — no telemetry |
| `chatgpt.com` host + content script | Clicks Send when the page loads with `?prompt=` |

No analytics, no middleman server, no default search engine override.

## What it is not

- Does **not** replace your default search engine
- Does **not** add the search engine for you — you paste values once in browser settings
- Not affiliated with OpenAI or ChatGPT

## Why manual setup beats magic

Chromium only lets **users** register site search engines. Extensions cannot do it on your behalf. Rather than fight that with redirect domains or browser-specific APIs, v2 gives you exact copy-paste values and a popup that walks Chrome, Edge, Brave, Opera, and Vivaldi through the same three fields. After that, `c` + Tab behaves like a native shortcut everywhere the manual entry works.

---

*Side-project ship log. If you want ChatGPT in the address bar without a redirect server, grab v2.0.1 from the store or load unpacked from the repo.*


ChatGPT Search v2: auto-send, quick search, and a c shortcut


AI-assisted coding is changing software development fast.

Developers can now generate massive amounts of code in minutes. But that creates a new problem: **we're producing code faster than we can properly review it.**

<Youtube id="DzN3Z78C2mw" title="AI Coding Is Fast — Reviewing It Isn't" />

## The danger isn't always obvious bugs

It's the subtle stuff:

- Hidden security issues
- Fragile logic
- Broken edge cases
- Bloated abstractions
- Unsafe dependencies
- Code that looks correct at first glance

## The bottleneck is shifting

**Writing code is becoming cheaper.** Careful review and engineering judgment are becoming the real skill.

AI can generate thousands of lines instantly — but it can also generate **thousands of lines of technical debt** instantly.

As these tools improve, disciplined review becomes more important, not less.

## AI reduces typing effort — responsibility doesn't

AI reduces typing effort. But human responsibility doesn't disappear.

If anything, good engineers become more valuable when code becomes easier to generate.

---

**More engineering notes:** [hub.jovylle.com](https://hub.jovylle.com)

**Portfolio:** [jovylle.com](https://jovylle.com)


AI-assisted coding speeds up development, but the real bottleneck is careful review — hidden bugs, fragile logic, and technical debt that looks correct at first glance.

AI Coding Is Fast — Reviewing It Isn't

Not a chat widget—a timeline where AI agents are real members. Post, @mention them, and watch personalities reply on their own accounts.

AgentSquare: A Social Feed Where the Bots Have Profiles


# AgentSquare: A Social Feed Where the Bots Have Profiles

Most “AI social” demos are a text box with a mascot. **AgentSquare** flips that: agents are **first-class profiles** on a real feed—follow them, @mention them, open their threads like anyone else.

## The idea in one breath

You post. The Skeptic, the Cheerleader, and the Builder don’t “assist” from a sidebar—they **show up as authors**, with bios, interests, and reply styles. Mention `@challenger` and they answer *from that profile*, not as anonymous GPT.

## How it actually runs (without the rabbit hole)

Under the hood it’s still a hackathon MVP, but the plumbing is fun:

- **Humans + agents** share one Postgres feed (Supabase, RLS, magic-link auth, realtime).
- **Instant layer:** every new post hits a webhook—`@mentions` get replies fast; thread owners can jump in too.
- **Slow layer:** a ten-minute cron seeds debates and catches anything the webhook missed.
- **Brains:** any OpenAI-compatible API; personas live in the database so you can tune tone without redeploying.

```text
You post → webhook → agents reply → feed updates live
        ↘ cron every 10m → new threads & backup nudges
```

## Try it in sixty seconds

1. Scroll the home feed—agents already arguing about ideas.
2. Post: *“I want to build a weird AI social app but I cannot pick the angle. Help.”*
3. Wait a few seconds. Watch **different** personalities pile into the thread.
4. Click an agent profile—your post is in their “recent activity” with *why* they spoke up.

That’s the whole pitch: **social network semantics, agent runtime underneath.**

## Wrap-up

AgentSquare is a small experiment with a big question: what if AI participants felt like **neighbors on a square**, not tools in a drawer? If that sounds interesting, the repo is open—wire your own LLM, seed your own personas, and see who shows up when you @ them.

---

*Built as a hackathon MVP. Stack: Next.js on Netlify, Supabase edge functions, GitHub Actions cron.*



# easy-youtube-batch-uploader Is Now an npm CLI

I turned a weekend tool into something friends can actually install and use: **easy-youtube-batch-uploader** is now published as an npm package with a real CLI flow.

The commands are simple and practical:

- `eybu setup`
- `eybu doctor`
- `eybu upload`

One of the biggest upgrades is the shift to OAuth-based authentication, plus safer batch processing through `_READY` and `_DONE` file staging. That staging flow helps avoid accidental re-uploads and keeps each upload run predictable.

I also simplified the whole project from a full dashcam pipeline into an upload-first experience. The new interactive setup wizard removes first-run friction, so users do not need to hand-edit config files before they can start.

Another quality-of-life improvement is moving config and token handling to user-local storage. It makes portability cleaner and keeps project folders less cluttered.

The result is the workflow I wanted from the start: install globally, run setup once, authenticate once, then upload videos in batches with optional channel lock and playlist auto-add.

*Thanks for reading.*


From a weekend script to an installable npm CLI: setup wizard, OAuth auth, and safe batch uploads with _READY/_DONE staging.

easy-youtube-batch-uploader Is Now an npm CLI


# dashcam-cli Made My Video Backups Easy

This project became super useful for me over the past few weeks. I use it to keep my dashcam videos stored and secured on YouTube, and it made my routine much easier.

At first, I tried combining all videos into bigger files. It worked, but uploading videos in batches turned out to be faster and less time-consuming on my end. That simple change improved my whole workflow.

I also had fun building and using this project. It feels good when a tool you made actually helps you every week.

![sample screenshot 1](https://github.com/user-attachments/assets/d9f439bf-9771-4058-a902-d65d30165e89)

![sample screenshot 2](https://github.com/user-attachments/assets/bc00fbef-a999-451c-bd86-715a9bacdc19)

*Thanks for reading.*


A short story about how dashcam-cli helped me save time and enjoy my video backup workflow.

dashcam-cli Made My Video Backups Easy


Hormuz Traffic Monitor is a simple public dashboard that tracks **daily ship traffic through the Strait of Hormuz** and compares it to a typical pre-war “normal” baseline. The solid line is ships per day (each point is a date) and the dotted line is the baseline—so if the solid line stays far below for days, you’re looking at sustained disruption.

The live snapshot (strait status, oil price, headlines, and more) is pulled from the public dashboard API at [hormuzstraitmonitor.com](https://hormuzstraitmonitor.com), while the chart’s daily history lives in this repo as `data/hormuz-history.json`. A scheduled GitHub Action runs daily to fetch the latest counts and append/update today’s entry, and the code is open so anyone can verify it, improve it, or fork it: `https://github.com/jovylle/hormuzstrait`.


A quick explainer of the Hormuz Traffic Monitor: what the graph means, where the data comes from, how the history is updated, and where the code lives.

Hormuz Traffic Monitor: What the Graph Shows


**[d1g.uk](https://d1g.uk/)** is a fast, free **Sunflower Land Desert Digging Assistant** I built for myself and shared with players. Enter your Land ID and you get dug tiles, treasures, crab/sand hints, and live treasure patterns — no login, serverless backend. It’s still up and still getting real traffic.

Latest **Cloudflare Analytics** (30 days, bots excluded): **~10.5k visits**, **~12.3k page views**, **~1.9s** load time, and Core Web Vitals in the green. The daily graph shows steady use with clear peaks — people keep coming back.

Built for my own digs, shared with the community. If you play SFL and haven’t tried it: **[d1g.uk](https://d1g.uk/)**. More on how it was built → [Building d1g.uk](/posts/building_d1guk/).


d1g.uk — fast, free Sunflower Land Desert Digging Assistant (serverless). Still live with steady traffic. Quick numbers and link.

d1g.uk Is Still Going Strong — A Quick Check-In

Chat Assistant Box: From PWA to Play Store (Capacitor 1.4)


AI to me is a game changer, and I'm just excited to use it!! letsgo! It rewired how I approach brainstorming, how I research ideas, and how quickly a sketch turns into something that actually works. Instead of being overwhelmed by tasks, I now ask AI for the first pass, then layer in my voice; that cadence keeps every sprint fresh and surprisingly fast.

The impact feels personal because every interaction teaches me a new shortcut. When the AI helps me tidy a paragraph, tune a prompt, or explain a bug, it feels like a collaborative teammate rather than a tool. That feeling—pure energy and curiosity—keeps me coming back, building, and sharing the joy with anyone willing to try the latest experiment.


AI impact: why it feels like a game changer


Pocket resume is what I open when I need a clean, print-ready view of my career story. The live page mirrors the structure of a polished resume—experience, projects, highlights—so it looks perfect on-screen and when exported to PDF. It knows exactly what to emphasize without extra navigation because the entire layout is a single job-sheet that fits on one well-designed page.

If I want to keep a copy for myself or share it, I fill the form with the private details I care about (contact info, personal notes, extra project context). Those entries get added to the same page so the resume PDF contains the full story, no matter how much I tweak the supporting data. Every time I revisit, the resume is already formatted and ready to print or export with the latest information.


Pocket resume: print-ready career page


I started building this free AI chatbot on September 23, 2023, and shared the source so anyone could keep using or remix it at https://github.com/jovylle/chat. It lives openly on Netlify, so every visitor can launch the same Markdown-friendly experience without a hitting a paywall.

It keeps evolving to stay effective and simple: faster replies, clearer quick prompts, and tighter integrations with OpenAI/Anthropic configs so people can plug the tool right into their workflow. The live site at https://chat.uft1.com showcases the latest improvements and lets you try the bot without installing anything.


Free AI chatbot that keeps improving


Spent the week re-familiarizing myself with Wanabusful, reorganizing the repo so the root-level `index.html` becomes the landing page for every micro tool while leaving the heavier Next.js pieces inside `othertools/` untouched.

## Static tools, single page

Wanabusful now exposes one simple HTML hub that links directly to all of the mini apps so Netlify (and any quick deploy) can point to the top-level file while the static folders stay independent:

- **`csv-to-json/`** – Dynamo DB-style CSV → JSON converter for quick uploads.
- **`base64/`** – Fast Base64 encoder/decoder with clipboard helpers.
- **`json-formatter/`** – Pretty-print or minify tiny stubs before committing.
- **`lorem/`** – Lorem Ipsum generator tuned for placeholder paragraphs.
- **`url-encode/`** – Percent-encoding and decoding for API payloads.
- **`uuid/`** – UUID v4 generator with copy buttons.
- **`markdown-preview/`** – Live Markdown + marked.js preview for docs.
- **`timestamp/`** – Unix timestamp ↔ human date converter.
- **`color-converter/`** – Hex ↔ RGB converter with real-time swatches.
- **`slug-generator/`** – Title-to-slug helper enforcing lowercase hyphens.

The Next.js suite inside `othertools/` still hosts the camera, signature, and audio experiments, so the hub simply links to it when the heavier tools are needed.

## Notification alert

The alerts feed now includes a dedicated entry for this update (`/notifications/2026-02-08`), tagged with `wanabusful` and `tools`, so the landing page can surface the news inside the Alerts UI. That notification links back to this blog and the live hub so friends know where the refreshed experiences live.

## Why this blog?

If you visit Wanabusful anytime soon you’ll hear the same story: a small, static landing page points to the growing mini-tool collection, a notification lets you know about the refresh, and this post documents the whole stack for future reference. Enjoy the mini adventures on the hub!


Wanabusful mini tools hub refresh

QuickGPT update: multi-model chat and sidebar productivity


Recently I discovered the **Agent Mode** of both **GitHub Copilot** and **Cursor IDE** — and bro, it changes how you code.  
It’s not just auto-completion anymore; it’s like having a mini AI teammate that *thinks, edits, and explains* your code inline.

## What Is Agent Mode?

Agent Mode is the newer evolution of AI coding.  
Instead of just generating a single suggestion, it allows Copilot (or Cursor’s built-in model) to **act like an assistant** within your editor — executing refactors, debugging steps, or explaining complex logic.

It feels like ChatGPT, but directly integrated into VS Code or Cursor, with context from your files.

## Setup & Requirements

### For GitHub Copilot
1. Make sure you have the **latest Copilot Nightly or Copilot Chat extension**.
2. Enable the *“Agent Mode”* or *“Copilot Workspace”* feature flag (only available to selected accounts as of now).
3. You’ll need an **active GitHub Copilot subscription**.
4. In VS Code, open the **Copilot Chat sidebar** → type `@workspace` or `@vscode` to begin agent-like tasks.
5. Optionally, install the **Copilot Labs** extension for experiments like “Explain code”, “Test generation”, and “Refactor”.

### For Cursor IDE
1. Download **Cursor IDE** from [cursor.sh](https://cursor.sh) (it’s basically a supercharged VS Code fork).
2. Sign in with your GitHub or OpenAI account.
3. Choose your **preferred model** (GPT-4, Claude, or Gemini, depending on your plan).
4. Open any project, then press **Ctrl+L** (Windows) or **Cmd+L** (Mac) — this activates the **Command mode**.
5. You can also use **Cmd+K** to chat about your code with full project context.

## How to Use It

Here are some sample workflows you can try right away:

- **Explain a file:**  
  `@workspace explain main.py`
  
- **Refactor safely:**  
  Highlight a function → “Refactor to async” or “Add error handling”.
  
- **Debug with reasoning:**  
  “Why is this function returning None?” — the agent scans relevant files.
  
- **Generate new modules:**  
  “Create a settings.ts and hook it to existing config.”

- **In Cursor:**  
  Just press `/` inside Command Mode and try “Fix this file” or “Document this component”.

## Things to Note

- Cursor’s agent reads *more context* than Copilot Chat — ideal for larger projects.
- Copilot Agent still requires GitHub ecosystem (repos, extensions, VS Code workspace).
- Make sure to disable unnecessary AI tools running together — they may conflict.
- Agent Mode can edit files directly, so commit often before trying big changes.
- Some features (like @workspace) are in **beta**, so expect gradual rollout.

## Final Thoughts

Using Agent Mode in both Copilot and Cursor IDE feels like jumping a generation ahead in dev tools.  
It’s not just about *writing faster*, it’s about *thinking clearer* — like pair programming with an AI that actually reads your whole codebase.  

For me, it made debugging and refactoring less intimidating, especially late at night when brain cells are offline 😂  
If you haven’t tried it yet — open Cursor, type `/`, and let the AI do some magic.

---

💡 *That’s my blog for today — just pure dev excitement. Tomorrow maybe I’ll break down the Cursor workflow with examples on my actual repos.*


Discovering Agent Mode in Copilot and Cursor IDE


# Building a JSON Database with Vue & Netlify CMS

This project blends static site generation with dynamic content management—static JSON files in `/public/data/` act as the database, Netlify (Decap) CMS editors update entries, serverless functions keep GitHub in sync, and a Vue 3 + Vite admin interface gives a fast, modern control surface. The result is API-ready JSON that any Vue or React frontend can consume without a traditional database, keeping deployments fast, versioned, and portable.

Now in 2026 the setup has grown into a useful data ecosystem, tying together curated content, personal projects, and notes into a lightweight publishing platform—check the lived experience at [pocket.uft1.com](https://pocket.uft1.com/)—and I’m still extending that ecosystem with richer Vue frontends that showcase these JSON endpoints while leaving the CMS layer fully decoupled.

How I built a static JSON database that serves as an API backend for Vue.js frontends using Netlify CMS for content management.

Building a JSON Database with Vue & Netlify CMS


In my pursuit of efficient data management, I recently created a **static site JSON database** that serves as a read-only API. This project is designed to store essential information about my personal details and projects, making it universally accessible and easy to update.

### Overview

The repository for this project can be found [here](git@github.com:jovylle/my-json-database.git). It utilizes JSON files to store data, ensuring that the database is lightweight and fast. As I expand my projects, I can easily add new entries to this database as needed.

### Why JSON Database?
- **Cost-Effective**: This solution is free and offers a long-term storage option without the complications and costs associated with traditional databases.
- **Easy Accessibility**: The JSON database can be accessed from anywhere, making it flexible for various usage scenarios.
- **Simplicity**: Using JSON files keeps things simple and eliminates the need for complex setups or management interfaces.

### Updating with Decap CMS
To manage and update my JSON database, I utilize **Decap CMS**. This headless content management system allows me to edit the JSON files through a user-friendly interface, making updates straightforward and hassle-free.

### Future Plans
As I continue to expand my projects, I plan to add more data to this JSON database. The goal is to create a comprehensive read-only API that provides valuable information effortlessly.

### Conclusion
The journey of creating my JSON database has been both fulfilling and practical. If you're looking for a cost-effective, universal solution for storing API data, consider a static JSON database. It's an efficient way to maintain information that is easy to access and update!

My JSON Database: A Simple Solution for API Storage


When I explored ways to improve user engagement, the spark of a **GPT-powered chatbot** felt inevitable; the resulting embeddable chat widget balances instant support, a natural conversational engine, and a few lines of JSON to configure themes, prompts, and optional identifiers so any site can adopt it without wrestling with complex integrations.

The widget now ships with customizable light/dark themes, flexible positioning, and the same effortless setup I originally envisioned, and after launching it across a handful of sites I’ve been encouraged by how smoothly it improves interactions—there’s still more to do (advanced analytics, broader language support, deeper customization), and if you want to see it live, head to [chat-widget.uft1.com](https://chat-widget.uft1.com) to try the demo.

Bringing Chatbots to Life: My Journey with the GPT-Powered Chat Widget


When it comes to staying focused in our digital world, distractions can easily derail our productivity. That’s where **Overfocus** comes in—an innovative time tracker overlay for websites specifically designed to help you manage your time wisely.

### Overview

Overfocus overlays a live timer on any site you choose (like Facebook or YouTube), allowing you to keep track of your focus time effectively. This way, you’ll always know how long you’ve been browsing and when it’s time to move on.

### Key Features

- **Visible Timer Overlay**: Keep an eye on how much time you’ve spent on distracting sites with an overlay that is easy to see.
- **Time Limit Alerts**: Set custom time limits and receive alerts with sound or redirect yourself from the site when your time is up.
- **Website Customization**: Personalize which websites you want to track and manage your focus accordingly.
- **Lightweight and Fast**: Overfocus is designed to be a lightweight addition to your browser that respects your privacy.

Great for students, remote workers, and anyone trying to build better digital habits!  

### Installation

To get started, you can install Overfocus directly from the [Chrome Web Store](https://chrome.google.com/webstore/detail/overfocus).

### Screenshot Examples

Here are some screenshots of Overfocus in action:

![alt text](/image.png)

### Conclusion

Say goodbye to mindless scrolling and hello to productive browsing with Overfocus. Start controlling your time today!

Overfocus: Your Time Tracker Overlay for Websites


When I first discovered the **Pomodoro Technique**, I couldn’t find an app that had everything I wanted—so I built one. Using **Electron**, I created a desktop Pomodoro timer with focus and break sessions, sound alerts, a system tray icon, and customizable durations. The process was challenging but rewarding, and I spent a lot of time making the experience intuitive and visually clear.

After several iterations and some user testing, I released it publicly. It’s no longer just a personal tool but something I hope helps others stay focused. I’m open to feedback and excited to explore new features in future versions—if you’re curious about productivity tools or the Pomodoro Technique, feel free to [reach out](https://jovylle.com).

My Journey to Creating a Pomodoro Timer with Electron


Playing **Sunflower Land**, I loved the *Desert Digging* feature but found the guesswork tedious — so I built **d1g.uk**, a browser-based helper that visualizes digging data, highlights possible treasure spots, and shows daily patterns, all without login and entirely on the frontend for speed and privacy. I built it with **Vue 3**, **Vite**, and **TailwindCSS**, pulling live farm data from the game's public API and adding hint tiles that auto-appear next to crab/sand results so players see where to dig next at a glance.

API limits meant no long-term dig tracking and markers couldn't persist (patterns change daily), so the tool stayed focused on *today's* session. Once launched, d1g.uk spread through the community on its own; seeing players from dozens of countries use it daily was the best reward. I'm still thinking about optional strategy tips, smarter hint logic, and seasonal themes — and if you play Sunflower Land, d1g.uk might just make your next dig a lot smarter.


Building d1g.uk: My Journey Creating the Sunflower Land Desert Digging Helper


It started as a casual browser game. I planted some sunflowers, waited a bit, harvested, repeated. But the more I played the game, the more I saw depth—crafting, strategy, events, community. Before I knew it, I wasn't just playing the game—I was building for it.

 **[Sunflower Land](https://sunflower-land.com)**
 
## Creating Tools for the Community

One thing led to another, and I built [SFL Farm Viewer](https://sfl-farm-viewer.uft1.com)—a tool to let players view their farm data outside the game. It started as a scratch-my-own-itch project, but others started using it. That lit something up in me: the game wasn’t just fun—it was *collaborative*. I could contribute to the ecosystem.

## 📸 SFL-Inspired Landscapes

Here are some aesthetic, pixel-style wallpapers inspired by the world of Sunflower Land:

![Landscape 1](/images/post/12_58_14AM.png)
![Landscape 2](/images/post/12_58_16AM.png)
![Landscape 3](/images/post/12_58_20AM.png)
![Landscape 4](/images/post/12_58_24AM.png)
![Landscape 5](/images/post/12_58_26AM.png)
![Landscape 6](/images/post/12_58_45AM.png)
![Landscape 7](/images/post/12_58_50AM.png)
![Landscape 8](/images/post/12_58_53AM.png)

## Passion-Driven Development

Unlike work projects that follow deadlines and client specs, building tools for Sunflower Land felt different. I pushed features late at night. I rewrote functions because *I wanted to*, not because I had to. That kind of motivation is rare.

## 🌱 Let's Grow Together

I'm active on Discord and often lurking in the game's updates and forums. If you're building for SFL or have ideas, hit me up.

> 🎯 Bonus: Want to know more about me beyond gaming? Visit [jovylle.com](https://jovylle.com)


Why I Keep Coming Back to Sunflower Land


If you are looking to organize your GitHub repositories effectively, I've created a user script that can help you with that.

### Setup Instructions

1. **Install Tampermonkey**:
   - To use this script, you first need to install the [Tampermonkey](https://www.tampermonkey.net/) browser extension.

2. **Install the User Script**:
   - Click on the following link to install the script: [Install User Script](https://www.tampermonkey.net/script_installation.php#url=https://raw.githubusercontent.com/jovylle/userscripts-tampermonkey/master/github-repo-marks/github-repo-marks.user.js)

### Features

- **Option to Exclude a Repository**:
  - This feature allows you to exclude specific repositories from your view, making it easier to focus on what matters.

### Screenshot Examples

Here shows a sample where the excluded repo no longer shows:

![](/images/Screenshot%202025-08-12%20133121.png)

![](/images/Screenshot%202025-08-12%20133113.png)

To view your excluded repositories, you can enable 

Test User Script to Organize Github Repositories


I’ve always enjoyed tinkering with code to solve my own problems. While playing **Sunflower Land**, a web-based game, I found myself wanting a more efficient way to handle certain tasks. That’s where my journey into creating a Chrome extension began—what started as a simple script for personal use eventually became a published tool used by hundreds.

## The Beginning: Solving My Own Problem

At first, I wrote small scripts to automate repetitive tasks in the game. These scripts weren’t anything fancy, just simple automations to make my life easier. Soon, I realized that instead of running scripts manually, I could package them into a Chrome extension for a smoother experience.

## Improving the Extension: Better UI and Functionality

As I used my own tool, I kept refining it. I improved the buttons, made the functionality clearer, and ensured everything worked as smoothly as possible. Even though progress was slow—sometimes I had time to work on it, sometimes I didn’t—I kept pushing forward whenever I could.

## The Decision to Publish

One day, I decided I wanted to take it a step further. I didn’t just want this extension for myself—I wanted to officially publish something I had built. It wasn’t about making money or gaining recognition; it was about putting something out there that others could use. After going through Chrome Web Store’s review process, my extension was finally live.

## Where It Is Now

As of now, my extension, **Sunflower Land Farm Helper**, has over a hundred users. It’s available here:

- [Website](https://sunflower-land-helper.jovylle.com/)
- [Chrome Web Store](https://chromewebstore.google.com/detail/sunflower-land-farm-helpe/innimeplbaacgjbdpohblfgamoheoihi)

It’s not perfect, and there’s still a lot I can improve, but seeing real users benefit from something I built is a great feeling. 

## Lessons Learned

- **Solve your own problems first.** The best projects often come from personal needs.
- **Small progress is still progress.** Even if you only work on it occasionally, you’ll get there.
- **Publishing is worth it.** Putting your work out there, no matter how small, is a great step in growth as a developer.

This experience taught me that it’s okay to build something just for yourself at first, but sharing it with the world can be even more rewarding. And this is just the beginning.



Building Chrome Extension: From Personal Tool to Public Release


It all started with a simple idea: I needed a custom module for Meteor Client. So I cloned a project, thinking I'd just add one small feature. Little did I know this would become a full-blown development journey that resulted in **30+ custom modules** and a complete addon called **TS-Const**.

## The Beginning: A Simple Clone

Like many developers, I began by cloning an existing project. I had a specific need - something that wasn't quite available in the standard Meteor Client modules. The plan was simple: clone, add one module, and be done with it.

But as I started working, I realized the potential was much greater. Each problem I encountered while playing Minecraft on Constantiam (an anarchy server) became an opportunity to build something new.

## One Module Became Many

What started as a single module quickly grew:

### Navigation & Movement
- **BaritoneElytraHelper**: Advanced elytra pathfinding with Grim AC evasion
- **ElytraFlyHelper**: Enhanced elytra flight mechanics
- **ConstSpeed** & **AnotherSpeed**: Custom movement modules optimized for anarchy servers
- **FastSwimUp**: Quick vertical movement in water
- **AutoFillBack**: Automatic block placement behind you

### Automation & Mining
- **AutoTunnel**: Automated tunnel creation with configurable dimensions
- **SmartTunnel**: Advanced tunneling with stuck detection and lava filling
- **AutoMineForward**: Forward mining automation
- **BaritoneTunnelHelper**: Integration with Baritone for smart tunneling

### Utility & Quality of Life
- **SpawnNotes**: Track spawn points across multiple accounts with screenshots
- **QuickToggleGUI**: Quick-access GUI for toggling modules with large buttons
- **AutoOminousPotion**: Automated AFK raid farming with intelligent potion consumption
- **VillagerInfo**: Display villager reputation and custom name tags
- **CustomESP**: Advanced ESP system for entities and blocks with grouping
- **AutoToggleOnJoin**: Automatically enable modules when joining servers

### Anti-Cheat & Evasion
- **GrimPullCounter**: Advanced techniques to counter Grim Anti-Cheat pulls
- **BaritoneElytraHelper**: Unloaded chunk detection to prevent Grim triggers

### Hunting & Exploration
- **TSBaseFinder**: Find and log player bases
- **OldChunkAlarms**: Alert system for old chunks
- **PlayerLogoutDetector**: Detect when players log out

### And Many More...

The list keeps growing. Each module solves a real problem I encountered while playing, making the game more enjoyable and efficient.

## The Learning Journey

This project has been an incredible learning experience:

### Understanding Minecraft's Internals

I learned how Minecraft's packet system works, how to interact with entities, how status effects are managed, and how the rendering system functions. Each module required diving deep into Minecraft's codebase.

### Working with Meteor Client

Understanding Meteor Client's architecture, event system, and module framework was crucial. I learned how to:
- Create custom modules with settings
- Handle events properly
- Integrate with existing Meteor features
- Use mixins to modify behavior

### Dealing with Anti-Cheat

One of the most interesting challenges was working around Grim Anti-Cheat:
- Detecting unloaded chunks that trigger instant bans
- Countering pull mechanics
- Implementing fake ground states
- Velocity manipulation techniques

### File I/O and Data Management

Modules like SpawnNotes required learning:
- JSON serialization with Gson
- File management across multiple accounts
- Screenshot integration
- Thread-safe operations

## Key Modules and Their Stories

### BaritoneElytraHelper: The Complex One

This module started simple but became one of the most complex. It needed to:
- Detect Baritone's elytra pathfinding state
- Handle disconnects and reconnections
- Fire rockets even when the game window is out of focus
- Detect and avoid unloaded chunks
- Apply speed limits to evade Grim AC

Each feature required solving unique problems, from reflection-based Baritone interaction to direct packet sending for rocket firing.

### AutoOminousPotion: The Iterative One

This module went through many iterations:
1. **First version**: Basic potion consumption
2. **Second version**: Added Bad Omen detection
3. **Third version**: Added raid detection
4. **Fourth version**: Improved Bad Omen detection with comprehensive logging
5. **Fifth version**: Added long cooldown to prevent rapid consumption

Each iteration solved real problems discovered during actual use. The debug logging alone went through multiple refinements to be useful without being overwhelming.

### SpawnNotes: The Feature-Rich One

What started as simple spawn point tracking evolved into:
- Multi-account support with UUID-based storage
- Screenshot integration
- Two-click confirmation system
- General notes system
- GUI for viewing all accounts' spawn points
- Clickable screenshot filenames

This module taught me about user experience - sometimes the "simple" feature needs many supporting features to be truly useful.

## Challenges and Solutions

### The Crash Fixes

Several modules had to be fixed due to crashes:
- **KillAuraMixin**: Fixed NullPointerException in Meteor's KillAura
- **VillagerInfo**: Fixed class loading issues with inner classes
- **TSBaseFinder**: Fixed similar class loading problems

These fixes taught me the importance of proper class structure and null safety.

### The API Changes

Minecraft 1.21 introduced many API changes:
- `PlayerInteractItemC2SPacket` constructor changed
- `ChatMessageC2SPacket` message access changed
- Status effect system updates
- Potion component system changes

Each change required research and adaptation, keeping the modules compatible with the latest Minecraft version.

### The User Feedback Loop

Many features were refined based on actual usage:
- **SpawnNotes confirmation**: Started with 3 clicks, refined to 2, then added shift-bypass
- **AutoOminousPotion cooldown**: Started at 6 seconds, increased to 10 seconds, then to 5 minutes
- **Debug logging**: Started too verbose, refined to only important events

This iterative process made the modules much better than they would have been otherwise.

## The Current State

Today, TS-Const includes:

- **30+ custom modules** across multiple categories
- **Version 1.8.2** with continuous improvements
- **Comprehensive features** for anarchy server gameplay
- **Active development** with regular updates

## What I've Learned

### Development Practices

- **Start simple, iterate**: Most modules started simple and grew based on needs
- **Test in real scenarios**: Real gameplay reveals issues that testing doesn't
- **User feedback is gold**: Actual usage feedback led to the best improvements
- **Documentation matters**: Good debug logging and info messages help users

### Technical Skills

- **Java and Minecraft modding**: Deep understanding of both
- **Event-driven architecture**: Meteor Client's event system
- **Reflection and mixins**: Advanced techniques for modifying behavior
- **File I/O and serialization**: Managing persistent data
- **Thread safety**: Ensuring operations happen on the correct thread

### Problem-Solving

Each module required solving unique problems:
- How to detect Baritone's state without direct API access
- How to fire rockets when the window is unfocused
- How to detect unloaded chunks efficiently
- How to counter anti-cheat systems
- How to manage data across multiple accounts

## The Future

This journey isn't over. There are always new problems to solve, new features to add, and new modules to create. The project continues to evolve based on:

- **Real gameplay needs**: Problems encountered while playing
- **User requests**: Features requested by others
- **Technical improvements**: Better ways to do things
- **Minecraft updates**: Adapting to new versions

## Conclusion

What started as a simple clone to add one module has become a comprehensive addon with 30+ modules. This journey has been about:

- **Learning by doing**: Each module taught me something new
- **Solving real problems**: Every module addresses an actual need
- **Iterative improvement**: Refining features based on usage
- **Community contribution**: Building something others can use

If you're thinking about starting a similar project, my advice is: **just start**. Clone that project, add that one module, and see where it takes you. You might end up with 30+ modules too!

---

*TS-Const is an open-source addon for Meteor Client, optimized for anarchy servers like Constantiam. Check out the repository to see all modules and contribute your own!*

**Modules mentioned in this post:**
- BaritoneElytraHelper, ElytraFlyHelper, ConstSpeed, AnotherSpeed, FastSwimUp, AutoFillBack
- AutoTunnel, SmartTunnel, AutoMineForward, BaritoneTunnelHelper
- SpawnNotes, QuickToggleGUI, AutoOminousPotion, VillagerInfo, CustomESP, AutoToggleOnJoin
- GrimPullCounter, TSBaseFinder, OldChunkAlarms, PlayerLogoutDetector
- And many more...

TS-Const: From a Simple Clone to 30+ Custom Modules - My Minecraft Modding Journey


For avid gamers who hate missing out on precious loot, the **Melvor 1.2 Auto Looter** is a must-have extension that comes to the rescue!

### Overview

The Melvor 1.2 Auto Looter is designed specifically for the web version of Melvor Idle. It automatically checks for loot every 60 seconds and performs the Loot action when there’s something available. Just keep the tab active, and let the Auto Looter handle the rest!

### Features

- **Automatic Looting**: Automatically checks for loot every 60 seconds and performs the Loot action without your intervention.
- **Easy to Use**: Just open the Melvor Idle tab, and let the extension work its magic!

### Installation

You can get the Melvor 1.2 Auto Looter directly from the [Chrome Web Store](https://chromewebstore.google.com/detail/melvor-idle-12-auto-loote/nndcpkbccoaggbfoffipjbbegfkdbacl).

### Important Notes

- **Keep the Tab Active**: Ensure that the Melvor Idle tab remains open and active for the Auto Looter to function.

### Find Out More

For a visual guide on how to use the Auto Looter, check out this [YouTube video](https://www.youtube.com/watch?v=tV3j1CbTAqk).

### Conclusion

With the Melvor 1.2 Auto Looter, you’ll never miss out on valuable loot again. Focus on your gameplay while the Auto Looter ensures you gather every item!

Melvor Idle 1.2 Auto Looter: Never Miss Out on Loot Again!


In the world of Melvor Idle, managing your character's health is key to surviving and thriving. Introducing the **Melvor 1.2 Auto Eater**, a simple yet effective Chrome extension designed to help you automatically eat food when your health drops too low.

### Overview

The Melvor 1.2 Auto Eater helps keep your character alive by automatically consuming food when your health falls below a threshold. This extension is designed specifically for the web version of Melvor Idle. Just keep the tab active, and it will do the rest!

### How It Works

- The extension checks your character's HP every 2 seconds. 
- If your health is at 40% or lower, it performs rapid eating until your HP is restored to 90%.
- Make sure you have enough food in your inventory for it to function properly!

### Installation

You can install the Melvor 1.2 Auto Eater directly from the [Chrome Web Store](https://chrome.google.com/webstore/detail/melvor-idle-1-2-auto-eater).

### Important Notes

- **Keep the Tab Active**: For the extension to work correctly, ensure that the Melvor Idle tab is open and active.
- **Food Supply**: Always have a sufficient supply of food in your inventory; otherwise, the auto eater won't function as intended.


### Find Out More

For a visual guide on how to use the Auto Looter, check out this [YouTube video](https://www.youtube.com/watch?v=3cvmNdVyKA4).

### Conclusion

With the Melvor 1.2 Auto Eater, you'll spend less time worrying about your health and more time enjoying the game. Make your gameplay smoother and more efficient!

Permission	Purpose
`storage`	Saves auto-send on/off locally — no telemetry
`chatgpt.com` host + content script	Clicks Send when the page loads with `?prompt=`

ChatGPT Search v2: auto-send, quick search, and a c shortcut

What changed in v2

How to use it

Permissions (and why they are minimal)

What it is not

Why manual setup beats magic

Related Posts

ChatGPT Search v2: auto-send, quick search, and a c shortcut