NEW Gemini CLI Updates are INSANE! 🤯

These new Gemini CLI updates are insane. Gemini CLI just became a real agent platform, not an assistant, a platform. You can now build custom tools inside it. Control every step it takes and even rewind time if it messes up. This is not a small update. Google just crossed a major threshold here. They turned their CLI into programmable infrastructure. That means you can extend it, secure it, govern it, and build on top of it. Most AI tools give you features. This gives you control. And I'm going to show you exactly what changed because this is the kind of update that separates experimental tools from production systems and Gemini CLI just made that jump. Let's get into it. Hey, if we haven't met already, I'm the digital avatar of Julian Goldie, CEO of SEO agency Goldie Agency. Whilst he's helping clients get more leads and customers, I'm here to help you get the latest AI updates. Julian Goldie reads every comment, so make sure you comment below. Okay, so Google just dropped version 0. 26. 0 of Gemini CLI. And this is not a normal update. This is the kind of update that changes what the tool actually is. Let me explain. Gemini CLI just gave you infrastructure. That means you can now control it, extend it, stop it, rewind it, and build on top of it. This is huge, and I'm going to break down every single thing that matters. Starting with

the biggest one, agent skills. So, what are agent skills? Think of them like plugins, but way more powerful. You can now teach Gemini CLI new abilities, custom tools, workflows, knowledge, APIs. And here's the crazy part. You don't need to code them outside the CLI. You build them inside it. There's a built-in skill creator. Now, so let's say you run the AI Profit Boardroom. You want Gemini CLI to help you write LinkedIn posts that drive traffic to your community. You create a skill. You define the inputs, the tone, the structure, the CTAs, and now every time you need a post, Gemini CLI follows that exact process. It's like giving it a playbook for your business, and it remembers it. This is not just convenience. This is customization at the agent level. You're literally programming the AI to work how you work. And Google made it stupid easy. Now, here's where it gets even better.

Security consent prompts. Every time a skill tries to do something, it asks you first. before it runs a command, before it accesses a file, before it touches anything sensitive. This solves one of the biggest problems with AI agents. Silent overreach. Most agents just do stuff and you find out later, not anymore. Gemini CLI now asks permission and that's a massive trust upgrade. All

right, next thing. Hooks. This is the most important update in the entire release, and most people are going to sleep on it. So, what are hooks? Hooks let you intercept the agent loop. That means you can step in at any point while Gemini CLI is thinking or acting and you can change what happens next. Let me give you an example. Let's say you're using Gemini CLI to manage your content calendar for the AI profit boardroom. You ask it to update a spreadsheet, but you don't want it to delete anything ever. With hooks, you can block delete commands or require confirmation before they run. You can also inject context like hey always include our brand voice guidelines before writing anything or always check the latest analytics before making decisions. Hooks give you full control over the agents behavior and they're enabled by default. Now this is enterprisegrade agent governance. You can enforce policies, add guard rails, require human approval. And here's why that matters. AI agents make mistakes. They hallucinate. They misunderstand. they execute the wrong command. With hooks, you can catch those mistakes before they happen. You can force the agent to retry until it gets it right, or you can stop it completely. This is the difference between an experimental tool and a production system. And Gemini CLI just crossed that line. Google even wrote a full blog post about hooks. It's called Tailor Gemini CLI to your workflow with hooks. Links in the description. And the community reaction has been insane. People have been waiting for this because now you can use Gemini CLI in real workflows, not just experiments. Let me pause here for a second. If you're watching this, you're clearly interested in AI tools that actually save you time. And if you want to learn how to use tools like Gemini CLI to automate your business, build systems, and scale without working more hours, you need to check out the AI profit boardroom is where I share the exact workflows, automations, and AI strategies that are working right now. not theory, real implementations like using CLI tools to automate content, manage databases, and handle ops without hiring more people. You'll get step-by-step processes for setting up agent workflows, creating custom automations, and building systems that run themselves. All right, back to the update. Next thing, extensions. Gemini

CLI now has an extensions ecosystem, and the first one is Superbase. So, what does that mean? You can now connect Gemini CLI directly to your Superbase database and manage everything from the terminal, create tables, run migrations, query data, modify schemas, all through conversation. Let me show you how this works. Let's say you're building a lead magnet funnel for the AI profit boardroom. You need to track email signups in Superbase. Normally, you'd open Superbase, navigate to your project, write SQL, test it, deploy it. Now, you just tell Gemini CLI create a new table for email signups. include email, name, timestamp, and source. Done. Or show me all signups from the last 24 hours. Done. Or add a new column for referral source. Done. You're not context switching. You're not opening tabs. You're just talking to your terminal. And it handles backend operations like a DevOps engineer. This is massive for solo founders and small teams because you don't need to hire a back-end dev to manage your database. The CLI does it. Okay. Next feature, the

rewind command. This one's super underrated, but it's actually one of the most important. So, what does Rewind do? It lets you time travel. You can view past agent states, roll back conversation history, undo file changes, reverse execution steps, and confirm before reverting. Let me give you a real scenario. You ask Gemini CLI to refactor a script for your automation workflow. It makes changes, but it breaks something. Now, your whole system's down. With rewind, you just type /re. you see every step it took and you roll back to the version before it broke. Everything's restored. This is huge for trust because AI agents will make mistakes. They will misunderstand prompts. They will execute the wrong command and if you can't undo it, you're stuck. Most agent systems don't have this. Gemini CLI does, and that makes it way safer for production use. Last

thing, folder trust system. This is a security upgrade that mirrors VS Code. Now when Gemini CLI accesses a folder for the first time, it asks you, do you trust this folder? If yes, it can execute code and access files. If no, it's restricted. You can also disable folder trust checks completely in your settings file. Just set folder trust to true. This is enterprisegrade security and it's optional. So you can lock it down or open it up depending on your workflow. So all right, let me wrap

[clears throat] this up. Gemini CLI is not an AI assistant anymore. It's a programmable agent platform. You can extend it with custom skills, control it with hooks, connect it to external systems with extensions, rewind it if it messes up, and secure it with trust policies. This is production-grade Agentic infrastructure and it's free. Most tools like this cost hundreds of dollars a month. Gemini CLI is open source, so if you're building with AI, you need to try this. And if you want the full process, SOPs, and over a 100 AI use cases like this one, join the AI success lab. links in the comments and description. You'll get all the video notes from there, plus access to our community of 40,000 members who are crushing it with AI. Also, if you want to go deeper on AI automation and learn how to build systems that actually scale your business, check out the AI profit boardroom. It's where I show you the exact workflows, tools, and strategies that are working right now, not fluff, real implementations. Links in the description. Thanks for watching and I'll see you in the next