Claude Code Mods Change EVERYTHING!
AI News Today | Julian Goldie Podcast
October 3, 2026
Claude Code Mods Change EVERYTHING!
Speakers Julian Goldie
Julian Goldie (0:00)
So we have a brand new update from Claude. This is Claude Code Mods. So I'm gonna explain what it is, and also give you 10 examples of how you can use it. Now, if you're wondering what a Claude Code Mod is, or you've never heard of it before, it's a new update that just dropped, and it basically allows you to change how Claude Code looks and feels, and you can build your own plugins in real time. So for example here, we can say build a Claude Code Mod called Aquarium. The Aquarium command opens a pane beside the session and draws the session as a fish tank. The main conversation is large parent fish. Each subagent spawn hatches a small fish that swims off and fades out when the subagent finishes. And you can see that's going to start building this. So it's writing the mod files inside Claude here. So you can basically change how Claude looks and feels. You can build your own plugins directly into it. You can even share these plugins as well, which is pretty cool and super easy to set up. So let's talk through what it is and how it works, first of all. So basically, a mod lets you change how Claude code looks and works. You can add your own screens, buttons, shortcuts, or set rules for what Claude is allowed to do. And here's how it works. So you can see an example of it working in real time right here. It's writing it in TypeScript or JavaScript. And there's basically five words that you need to know first, right? So you got the tool call. Now, basically, when you're using this, any single action Claude takes on your computer, like reading a file, editing a file, or running a command, this is a tool call. Then you have an event. So this is a moment when something happens inside Claude. We're just going to enable this plugin, you can see where it will come back to events. So this is a moment when something happens inside Claude. So for example, you can say, okay, when Claude is about to edit a file, that's an event, do this. And then you just press Enter and it's an event. Hook is a bit of code that says, run me, so run this Claude code mod whenever this event happens. So a mod is often a collection of hooks put together. And then we have the plugin. So this is a package of add-ons for Claude code, and a mod is one kind of plugin. You can install it the same way. And then you also have context. So Claude's short-term memory for the conversation, and it fills up as you chat. Now some mods can actually show you how full your context is as you go along inside the conversation, which is super useful. Let me show you an example of that. So you can see here how it works step by step. So how does this fit in to Claude code? There's basically four ways to add things to Claude code, right? Three of them work from the outside, and the mod is the only one that actually works from the inside of Claude, which you can build as you go along, which is why you can change the screen and the step into Claude's actions. So you got skills. Skills are like a note file that Claude reads. You got the MCP servers, and these connect Claude to other plugins and other tools outside. Then you also have hooks as well. But mods can basically live in your conversation as you go along. And that's the biggest difference right here.
Now, you can see this is building out the Aquarium mod, as you can see here.
So we can preview what it's building. And now we have the full Aquarium mod.
And from here, what we can do is we can type forward slash Aquarium. And now it's going to run this mod, as you can see. So that's an example of a basic mod. There's actually an amazing website that I found actually recently. Let me show you this.
And this is a website called mods.ai.dojo. And it has over 359 mods inside it that you can use, right? So these are all different examples of different mods that you can plug into Claw to Code, which is super useful. And you can switch between the categories here. So like writes and runs, or the network, or draws and remembers. It's got, for example, like fast compaction as well. So you can actually use GEV AI to create fast compaction in your conversation. Here's another example. So this is example number two. And you can see here that we can actually use a system here to measure the number of tokens we have. And this just shows up as a plugin on the side, right? So we've got a context window, the session cost, the plan limits, and this was built using the token meter as well. And you can literally describe anything that you want and then get Claude to build it out for you. There's three ways that this can show up, right? So if you want to indicate what can you actually do with this, in the morning, you can add your own screen. So you can add like a panel on the side or a strip of the typing box, buttons, tabs, text fields. You can also change what's already there. So you can restyle or replace parts of Claude code as well. You can also step in before Claude acts as well. And then you can make your own forward slash commands using mods as well. So for example, if I type in forward slash tokens now, we can easily call upon this context window tracker right here.
9 more minutes of transcript below
Thousands of transcripts fetched by people building searchable podcast archives
Fetch the whole transcript
The demo key returns a sample episode in full, no card needed:
curl -H "x-api-key: pt_demo" \
https://spoken.md/transcripts/1000651996090Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.
From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.
Using your own key:
curl -H "x-api-key: YOUR_KEY" \
https://spoken.md/transcripts/1000793016638