A unified AI coding workspace integrates multiple AI coding agents (such as Claude Code and Codex) into a single platform, enabling developers to manage agent sessions, files, tasks, diffs, commits, terminals, and diagrams in one organized environment while avoiding the resource-intensive limitations of running separate AI applications.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
The Missing Workspace for Claude Code and CodexAdded:
Code code is awesome, but trying to use it is a nightmare. It's so bad I moved to code X just to code properly. But now I've found the ultimate solution, Nimblicity. It allows me to talk to my agents, build diagrams. You can see here we're using chat GPT and Anthropic, different models, all in the same workspace for the same project using Nimblicity. Now Nimblicity is a completely free to use open-source harness for your AI agents. And if you're watching the channel, you'll know that I've been trying to build my own and it's an absolute nightmare. So the fact that there was an open-source one that we can just use off the bat amazing. So let me give you a quick run through of what this is, why you're going to use it, and then we'll get into actually using it itself. So first of all, it allows you now to have this agent mode where you can talk to different agents. So here's my Claude agent, as you can see here. If we pop over here, I have my chat GPT one.
And then I can open up another one and we can just start chatting here. And all of these sessions are then inside of a Kanban board. So I can come and see all the things that are happening on this project. And I can just pop open another one like here where I'm working on a new project. And over here, I can then work on this one. And so it keeps them separate. But the thing is, it doesn't kill your RAM like using code X or code code. Like these applications take up so much RAM, it's actually crazy. Whereas here in Nimblicity, it's so slick. And so you have these different chats which you can then use to edit files, etc. You can come and commit them with the AI here. Like if we go to this one here, it's implementing stuff. You can see the types of changes that it makes like with the diff. And then you can just simply click commit with AI and boom, it will then commit this and you can add a message, etc. Now not only can you see the agent view, but you also have a way to view all of the files as well. You can see here that it even creates like Excalidraw files which is amazing. Cuz if you know the channel, I come here and I have to go and open up my scalar draw all the time to then come and show you stuff. Well, there you go. I've already got some stuff in here. But instead of having to do this and then lose it and not know what project it's in, boom, it's all here directly inside Nimblelist. And also, your usage is so easy to see on the left. Like here look, Claude usage and Codex usage. And all of this is just like picked up because I'm logged in on my system. So, I don't have to come here and set anything up. It's amazing. So, let's open up this new one here and I'll actually show you how to get started properly for this. So, when you first run it, you're going to have to input your Claude code sessions. And then, you'll be able to get all the ones here that you've been using in Claude code. As I was saying, I'm using more Codex. But as I was saying, this project I've not been using recently. So, this is the only one we have. Now, what we can do then, just to get started, is just say something like, "We're just getting started again with this project.
What is the current status?" Then we can send that in. And now here you have your agent mode talking. And the cool thing about this agent mode is that it's super interactive. If it has questions, it's going to ask you them and pop them up in a nice slick UI way. And it also is like visualization first. So, if it can show you stuff like tables, for example, it will. Just like this research here, it creates a table to show me and a graph as well all inside of here. And now whilst this one is working, let me open up another one cuz I'm pretty sure the last time I was using this, I was messing around with the UI. So, let's open up instead Claude and we'll say, "Review the UI, please. Make sure the branding is consistent and slash plan any changes."
Boom.
And now you can see we have these two agents. One's Claude, one's OpenAI, which is amazing because I always go between them to try and get them to review things. And sometimes things like UI are maybe better for Anthropic rather than OpenAI. And if you try and use Claude saw, it's going to cost you an arm and a leg. Whereas Nimblelist here, completely free. Just uses your actual balance. And boom, there we go. So, it now has our output here and it is linking us to different files as well, which is pretty cool. And over on the top right, you can see it's in planning stage just automatically. And so, that means that then if we go to the Kanban, you'll see that these are now in planning. It's honestly amazing. And so then, the other one here, you can see that's kind of yellowy orange. That means that I need to go and check it out. And boom, as you can see, it's got questions baked directly into the UI.
It's so good. So, let's check this out here. We'll go auto only and then we'll go quick consistency pass. Let's check that later. It's a new update, which is cool because it's open source. They're always trying to work on stuff. So, if you have any feature ideas, like just come and request a feature just here. So anyway, whilst that one's going on, over here then, what we can do is we can say, "Visually show me using Excalidraw." And now it's going to go and use Excalidraw natively inside of here to show me what's going on. And if at any time I want to like change permissions, for example, over here, we have permissions.
So, we can change these. I'll leave allow all pretty much always, but you can change them if you want. You can even pop open a terminal, new terminal, and boom, we can start talking to it here as well. And it's directly inside of the folder that we're currently in.
So, now it's created this Markdown file.
We can actually check this out directly in the chat. Up at the top, just select on files, or in the right hand side, you can see here the uncommitted and changed files. And now we can see that the green is obviously new. And so, what we can do then from here is we can either revert, keep, etc. We'll just click keep. And then we can commit this with the AI over here. Commit with the AI, and it will send in a prompt down to the agent. So, if we select here, go back to the agent.
And of course, for all this clicking around, there are all shortcuts that you can use, but just clicking around so you can actually see what I'm doing. And there we go, we'll commit that. Boom.
And that's it, salted. Come back over to our other agent, see how he's getting on. He's still creating. That's fine.
But if you'll notice that they were in planning, well, what that does is over here, we can now go to the tracker and in the tracker, see we have this plan.
If we select it, you can see all the information. But basically, once this is done, the plan, it will then turn this plan into tasks and these tasks, you can then get AI to actually do. So, type hashtag task in markdown or use {slash} track in chat and you can just start using them. And so, we have our Excalidraw here done and we have this markdown file here. A better way to see those is to simply go over here to file mode and inside file mode, we now have our agent on the right-hand side and we can change between these different agents. And then on the left-hand side here, we can view our files. And so, if we open up this markdown one here, all we have to do is select it and boom, it opens it here on the left-hand side. And you can see that this is because we done a {slash} plan. You can actually see the plan details here with tags, priority, progress and the status as well. And then for the Excalidraw one here, if we click that open, you can see we have the canvas directly in here, which is crazy.
And if you change different like themes here, if I go to the light theme, then you'll actually be able to see properly like what it's doing as if you were on Excalidraw here. Also, a little tip, if you go to Monokai, that one keeps this in the the white mode, not dark mode. Um whereas the rest of the things go into dark mode, which is pretty cool. So, let me Excalidraw again and boom.
It's so good and it's actually creating them so well. There you go. Boom. We have our our current project here. And if we want to share this, all we have to do is right click, share link, then you can give it an expiry. So, how long do you want it to be live? Go 7 days, copy the link and then bam, you can share this with your teammate, whoever it is, they can come have a look. They don't even have to log in or anything as long as they have the link, then they're able to get it. And again, this is all part of the completely free Nimblelist.
Right, it's Luke from the future. I meant to say here, the cool thing about Excalidraw is that it's actually live as well, so So can edit it here. It's not just an Excalidraw file that you can view, it's an actual interactive Excalidraw board. So for example here, I have this file. I've done some stuff.
You can you know, write some text, whatever.
And then at the end right click close. Then if I go over here, can open it back up and bam, you can see that it's saved this and I can just come in, I can edit as normal as you would inside of a normal Excalidraw.
So now you never have to leave Nimblelist. And the last thing I'll show you today to show you just how amazing this is for actually talking to agents is inside here like for example, review UI branding. I can then just click plus and opens a new session inside of this like top level feature or chat and I can even change to like GPT 5.5 for example and I can say you know, review the review or you can have like different features under the same like major feature. So like little pieces. They can all be different ones here or you can ask questions, etc. But the fact that we can have multiple sessions from multiple AI models inside of one like master session is crazy. Especially if you compare this to like here for example, I won't open up but inside of Codex and my emails, I'm looking through emails all the time and one of the biggest problems I have is I need to keep creating new chats. But again, this is all only using GPT. Whereas if I want to use Claude, I'd have to then go and create the same thing and Claude and I've got multiple apps opened up and it's just it's so frustrating. But here we can see, GPT 5.5 is reviewing Claude's work to make sure that it looks good. And most models keep changing whether it's Codex or GPT this week is the best one, you can use the same harness. You can use Nimblelist for all of them. And so in short, I highly recommend Nimblelist. It is the tool that I've been missing and literally been trying to build myself with no luck for months. Check out the link in the description and download Nimblelist today. It's completely free and open source for individual use. So I'll see you in the next video. Take care. Bye-bye.
Related Videos
OpenHuman VS Hermes AI: Who Wins?
JulianGoldieSEO
285 views•2026-05-29
Long-Running Agents — Build an Agent That Never Forgets with Google ADK
suryakunju
142 views•2026-05-30
5 Mind Blowing Omni Uses Cases
PaulJLipsky
1K views•2026-06-02
This computer is made from real human brain cells. And you can buy it.
Talktmsmedia
3K views•2026-05-28
BREAKING: Microsoft’s New Image Generating Model Beat Out GPT 1.5 and Nano Banana 2
aimmediahouse
122 views•2026-06-03
I Made the Same Anime Fight Scene in Every AI Video Generator
NobleGooseAnime
295 views•2026-05-30
Nvidia Bets Big On AI PCs | New Chip To Power Windows Laptops | Technology | AI Updates | N18S
cnnnews18
3K views•2026-06-01
I Tested NEW Opus 4.8 on Four Projects (Updated LLM Leaderboard)
AICodingDaily
298 views•2026-05-29











