You are browsing as a guest. Sign up (or log in) to start making projects!

chahatesh

@chahatesh

Joined June 12th, 2026

  • 28Devlogs
  • 9Projects
  • 6Ships
  • 60Votes
Open comments for this post

1h 31m 32s logged

Devlog 7 – It Actually Works!
This was one of the biggest milestones for the project.The first major task was converting the app from a regular website into a full Electron desktop application. I’d never built an Electron app before, so this was completely new territory.

I spent a lot of time reading documentation, troubleshooting errors, and figuring out how everything fit together.

One of the first problems I ran into was my Node.js version. I was still using Node 20, but Electron required a newer version, so I upgraded everything to Node 22. After that, I finally got the app launching successfully.

Then I hit another challenge. The screen-sharing method I had been using in Chrome relied on Chrome’s built-in sharing APIs, but those don’t work the same way in Electron. Instead of trying to force it to work, I built a completely new system that can capture your screen directly inside the desktop app.

The Biggest Feature - AI Debugging This is by far the biggest feature I’ve added so far.

The app now works across your entire computer instead of being limited to a single browser tab. No matter what you’re working on, you can instantly ask the AI for help.

There are two shortcuts
Shift + Enter – Instantly captures your screen and sends it to the AI for analysis.
Ctrl + Shift + Enter – First asks what’s wrong so you can provide extra context, then captures your screen and sends both the screenshot and your description to the AI.

The best part is that everything gets added to the conversation’s context, so you can continue asking follow-up questions without having to explain everything again.This makes debugging incredibly fast. Instead of taking screenshots, uploading them, and re-explaining the problem every time, you just press a shortcut and the AI immediately understands what you’re looking at.

0
0
6
Ship

For this next ship, I mostly changed the overall design of the app. In the last review, it was said that the app looked a little too AI-generated, which is fair since I had used AI to help generate the CSS. So, I redid the CSS myself to make the app look much better. I am not the strongest at CSS, so it still has a slightly older look, but I think it turned out fine.

I also added a demo mode, which was an important improvement because one of the main pieces of feedback was that users had to make an account and sign in every time. With demo mode, they can try the app without needing an account. Another issue I worked on was the AI not always calling tools correctly. That took a lot of time to fix, but after a lot of work, I got it working properly. I also put more emphasis on voice typing, because that is one of the main features of the app. It makes everything faster and easier since users can just talk instead of typing everything out.

  • 3 devlogs
  • 4h
Try project → See source code →
Open comments for this post

1h 17m 52s logged

Devlog 8 - Ship 2
So based on last ship the biggest think was that people did not like the set up of the app and that it look bad so i made it diffrent and now I think it is better also i added a demo mode so you can test with out making a acount only change was that it was saved in local storage
BTW - if you are reading this this is a egg

0
0
2
Open comments for this post

3h 5m 9s logged

DevLog 6 - ReRevamp

So… my brother said my website looked bad. 😭

That was enough motivation to go back and redesign a bunch of the app’s CSS. The UI is much cleaner now, with better spacing, colors, and overall polish.

The coolest feature I added by far is screen shairing. At first, I tried adding an image upload system so the AI could check screenshots, but it ended up being too slow and interrupted the workflow. Then I realized Google Meet can access your screen directly through screen sharing, so I built something similar. Now, whenever you press Shift + Enter, the app captures your current screen, analyzes what’s happening, and explains what it sees.
It also automatically adds that information to the conversation context, giving the AI much more information when helping debug issues or answer questions.It’s way faster than constantly taking screenshots and uploading them manually, and it makes debugging feel much more seamless.

BTW I renamed the app to understand

4
0
35
Open comments for this post
Reposted by @chahatesh

3h 15m 58s logged

DevLog 5 - AI is stupid
So before the ai was very bad it did not look at the context and went on a limb so now I made the context better and the ai adds to it and it has links time and a bunch of other stuff so it is smart (Yay).

It can look at context and think becouse i swaped to a new ai - gpt-oss-120b - and it is smarter becouse more prameters

This took way too long and it was hard but i got it to work and i think this is good if anyone thinks there a probloms dm me or comment

1
1
25
Open comments for this post

3h 15m 58s logged

DevLog 5 - AI is stupid
So before the ai was very bad it did not look at the context and went on a limb so now I made the context better and the ai adds to it and it has links time and a bunch of other stuff so it is smart (Yay).

It can look at context and think becouse i swaped to a new ai - gpt-oss-120b - and it is smarter becouse more prameters

This took way too long and it was hard but i got it to work and i think this is good if anyone thinks there a probloms dm me or comment

1
1
25
Open comments for this post
Reposted by @chahatesh

1h 43m 21s logged

DevLog 4 – Cookies
I tackled one of the biggest issues with the app saving user data. My original design didn’t have any way to store progress, which was a pretty major flaw.My first attempt was using cookies. I spent quite a while implementing everything, but after I finished… nothing worked. I went through the code over and over trying to find the bug, but I couldn’t figure it out.

Eventually, I searched online and found the real problem I was testing on local host, cookies wasn’t going to work. Instead of fighting it any longer, I switched everything over to local storage.The change worked almost immediately, and it was a huge relief to finally see data being saved correctly(YAY)

I also spent some time working on voice typing. It looks like I have it working now, and I think it’s going to be one of the coolest features in the app. Being able to talk instead of type just feels like the natural direction for an AI-powered tracker.

PS in the vid I am reloading but you can’t see it

0
1
14
Open comments for this post

1h 43m 21s logged

DevLog 4 – Cookies
I tackled one of the biggest issues with the app saving user data. My original design didn’t have any way to store progress, which was a pretty major flaw.My first attempt was using cookies. I spent quite a while implementing everything, but after I finished… nothing worked. I went through the code over and over trying to find the bug, but I couldn’t figure it out.

Eventually, I searched online and found the real problem I was testing on local host, cookies wasn’t going to work. Instead of fighting it any longer, I switched everything over to local storage.The change worked almost immediately, and it was a huge relief to finally see data being saved correctly(YAY)

I also spent some time working on voice typing. It looks like I have it working now, and I think it’s going to be one of the coolest features in the app. Being able to talk instead of type just feels like the natural direction for an AI-powered tracker.

PS in the vid I am reloading but you can’t see it

0
1
14
Open comments for this post

1h 27m 54s logged

DevLog 3 - Revamp
This update was all about improving the experience. I switched the color scheme from blue to an amber theme after a friend pointed out that I might be using blue a little too much. Honestly, they had a point;(
I also expanded the notes system. Now, each work session keeps track of your timer, notes, and any relevant links, all in one place. The AI can summarize what you worked on, making it much easier to look back on your progress without reading through every note.

0
0
12
Open comments for this post

37m 42s logged

DevLog 2
I started working on the AI, and there’s now a built-in AI chat. At first it might seem a little random, but it’s actually a big part of what I have planned.The goal is to add a project journal where you can document your progress over time and easily share it with friends or anyone else who’s interested.
The AI will help organize your updates, answer questions about your project, and eventually make keeping a journal much easier.It’s still early, but this is one of the core features that will tie everything together.

0
0
16
Open comments for this post
Reposted by @chahatesh

3h 30m logged

Devlog 1
Spent a lot of time getting the foundation of the project up and running. The idea is to make a simple AI-powered project planner where you can keep all of your projects in one place without digging through a million notes.Right now, the app lets you create multiple projects, track work sessions with a built-in timer, and save notes for each session. Those notes are automatically organized into a running project log, so everything stays in one place instead of disappearing into the void.The AI is starting to come together too. It can summarize your progress automatically, and you can use voice input so you don’t have to stop what you’re doing just to type everything out.There’s still a lot to build, and I’m not 100% sure what feature comes next, but having the core working feels like a huge step forward. It’s not flashy yet, but every app starts somewhere… even if that “somewhere” is a timer and an unhealthy amount of debugging.

0
1
19
Open comments for this post

3h 30m logged

Devlog 1
Spent a lot of time getting the foundation of the project up and running. The idea is to make a simple AI-powered project planner where you can keep all of your projects in one place without digging through a million notes.Right now, the app lets you create multiple projects, track work sessions with a built-in timer, and save notes for each session. Those notes are automatically organized into a running project log, so everything stays in one place instead of disappearing into the void.The AI is starting to come together too. It can summarize your progress automatically, and you can use voice input so you don’t have to stop what you’re doing just to type everything out.There’s still a lot to build, and I’m not 100% sure what feature comes next, but having the core working feels like a huge step forward. It’s not flashy yet, but every app starts somewhere… even if that “somewhere” is a timer and an unhealthy amount of debugging.

0
1
19
Open comments for this post

1h 14m 10s logged

Dev Log 4
I added a Notes app, Music app, Settings app, and Weather app to my WebOS. They were all simple to build, and I chose “Life Goes On” by Oliver Tree for the music app because I like the song.

0
0
32
Open comments for this post

23m 48s logged

Dev Log 1
I made some comands and some api calls for those and made it so that the bot sometimes tells you what it is and some time does not and also i got it running on nest 24/7

0
0
3
Ship

What did you make
I made WebOS, a browser-based operating system with a Projects app, a Calculator, and a Browser.

What was challenging
Getting all the apps to work together smoothly and making the interface feel like a real operating system and the deployment was hard

What are you proud of
I'm proud that robo OS feels polished and includes multiple working apps and a hidden Easter egg.

Try project → See source code →
Open comments for this post

1h 11m 56s logged

Dev Log 3 - Done (Yay)

Added two new apps to WebOS: a simple calculator and a browser (powered by an iframe cool:) I also hid a Easter egg and polished the overall UI and i made github read me and also I have it ready to ship

0
0
4
Loading more…

Followers

Loading…