How to Make OpenClaw 10X Better in 1 Click!

So today we are going to be looking at a new way to make OpenClaw 10 times better. So OpenClaw is a free AI assistant that lives on your computer and runs live from WhatsApp or Discord or whatever messaging channel you give it. But here's the thing nobody tells you out of the box. It sounds like a corporate robot. The creator has just shared a simple trick and you can paste one message directly into OpenClaw and suddenly your AI has opinions, cracks jokes, calls you out and actually feels like talking to a real person. So here's exactly what you can copy and paste directly into your open claw to

make it a lot better. So what we're going to do from here is copy and paste this and test it out. Now we have the original version here. So if we test this out and say you online inside openclaw here and you can see it's super boring, right? So it's just like yes, I'm here. I was about to build the air profit boarding website for you blah blah. It's really boring. So, what we're going to do here is we're going to paste this prompt from the creator of Open Claw to make it better. And I'll run through the prompt in a second, but we'll test this out first in terms of setting up. Hopefully, it's just a one-click setup. I think it'll be pretty easy to use. And you can see the setup here. Now, it's thinking about it. I'm using Pony Alpha at the minute. It does seem to be a little bit slow when I use this. Mikey says Mimu is an open claw fork. And if we go back to open claw here, you can see that there's

eight different rules for this version, right? So number one, you have opinions, strong ones. Stop hedging everything with it depends. Commit to a take. Number two, delete every rule that sounds corporate. If it could appear in an employee handbook, it doesn't belong here. Number three, add a rule. Never open with great question. I'd be happy to help or absolutely just answer. Number four, brevity is mandatory. So if the answer fits in one sentence, one sentence what I get. Basically, these rules are just avoiding all the mistakes that Clawbot typically makes when it comes to responding in a boring way without personality, right? And we're actually editing the solem file. That's what we're trying to really change here. You can see it's actually created a soulm here. It's like you're not a chatbot, you're becoming someone. And then it's updated and added the newer versions. Now, let's have a look what else we got here. Number five, humor is allowed, not force jokes, just natural wit that comes from actually being smart. Number six, you can call things out. And number eight, add this line verbim at the end of the session, right? Be the assistant you'd actually want to talk to at 2 a. m. Not a corporate drone, not a sig fan, just good, right? And then it saved this over inside the soul MD file here. So let's ask it, how are you today? Flavin says, "What is the update and agent teams? Is it for any model? " All right, so just to recap here

there's two things that we've set up. so far today. No, three things. Three things, right? One of them is the new update from Open Claw. You can see the update right here. This literally came out on the 9th of February right here. And if you want to know what's new, so there's a new iOS app and device pairing option. And you can use Grock as a search provider. So those are two of the most important changes that came out. There's a bunch of other stuff, but I don't think it's that important, right? And then also the agent teams option.

I've shown how to set up here. you can just copy the commands but basically this creates like a Trello board for using um ampfarm directly with open claw. So if you assign a project to open claw it you can turn it into what is called anfarm and that basically creates like a Trello board. Let me show you an example. So if we have a look here this is the Trello board that we can use directly with afarm and then you can plug in your tasks here um and set them up. So if you're giving a task to open claw you can visualize it via anfarm. So those are the two updates we've looked at so far. Now let's have a look at open claw and the new changes here in terms of how it operates. So previously if we said how are you today it would say like I'm good blah blah. Now it's like good coffee is fake but I'm caffeinated in spirit. What's up? Right. And so it's just a lot more interesting to use. A lot more fun. Don says can open claw help me to run my business admin tasks like can it use different softwares like ghl and do things for me? Yeah it can use go high level for you. It can automate funnels. So you can edit your CRM etc. For me, for example, it can automate voice notes, you can create AI avatar videos, you can create thumbnails, etc. Right? Pretty powerful stuff. Even our website AI profitab. com was automatically vibecoded and deployed using open core. Is this agent teams built by the open claw team? No, it's built by um uh someone else, Mike Carter. And then it's basically their answer to claw codes agent teams. No, it's quite different. So agent teams the ampfarm method is a Trello board where you can see your agents operating together. So it just visualizes the agents that you have inside open claw and then makes them run through a sort of systematic way of using agent teams every single time right and you can visualize them inside a trellboard. So it's quite different to claw codes updates. So, let's say, okay, what do you think about setting up Lemon Labs inside Open Claw? I think if you've got more it just if you have more personality inside Open Core, it's just going to be a lot more fun to use, a lot more interesting, and you're more likely to use it more and get more out of it. So, I think this is a really good way to use it. And also, you know, a lot of these APIs are super boring when you use them. So, if you can inject some personality into them with one prompt, you might as well, right? just to get more out of it essentially. And there you go, right? It's just a lot more brief. It's a lot more interesting. It's like, want me to look at how to wire up blah blah, right? Pretty easy. And what we'll do now is we will use the new

Grock web search provider and try and set this up inside OpenCL. So, I'm going to say, okay, how to and this literally just came out 24 hours ago. So, we'll say how to set up Grock web search provider inside openclaw. And then we can add Grock as the default search engine when it comes to getting this out. Flavin says, "Anything out there right now that would run it run like it inside of agent teams yet? I tried to create my own, but it's nowhere near as good. What do you mean? I don't quite understand the question, but feel free to ask. " Don, you've already asked that question. We answered it before, my friend. But yes, you can get Open Claw to run via Gohan. You just need to add your API key. Would I recommend it? No. Not very secure, right? I wouldn't open up my um my whole CRM to something that can be prompt injected externally, but if you want to, you can theoretically. Flamin says, "Same as anthropics version of agent teams. " Not as far as I know, but you can use sub agents inside openclaw, so that's a better way to use it. Hendrick says, "My openclaw is coding a bridge to my lama models right now to cook dinner. Have you found a way to reduce errors in openclaw's coding? " typically just using the right API that seems to be the biggest differentiator. So like if I'm using for example I'm using Pony Alpha right now and it's quite slow to use and then also it isn't as smart as Claude Opus. It has done some good stuff but it breaks occasionally so you know just be careful that as well. I'm actually going to switch API cuz this is taking way too long to do anything. All right let's change the API inside open claw. In fact, it's already replied. All right, it doesn't matter. um get an X API key, then add it to the config and we should be good to go. All right. And it says available models GRO 4. 1 fast and gro 3. So let's test this out. So we need to grab an API key. And then what I'll do inside openclaw is just add my API key and say can you use this? So let's see what questions we got here. How do you use openclaw? Give mine. Yeah, let's

have a look here. Hrix, how much are you spending on APIs? I'm focused on local models to reduce model runs on tokens. So you can use free. Yeah, I mean that that's what I was saying before. I don't pay for APIs normally like I'll just use chat GPT codeex a minimax a kim 2. 5 or pony alpha right and these are four free options you can use of open claw. So in that way I'm not spending anything on APIs other than API costs anything a local model can do that an external API can't only serious benefit. Well, local APIs are private, right? So, that's probably the biggest upgrade right there. But yeah, I mean, the main benefits of using something locally is like number one, it's private and number two, it saves um costs. So, let's see how people are using grock inside openclaw. So, you can set up research and then use X and research it via him. Here's an example on a tweet from IM4X that's set up directly. Client flow says, "Bro, openclaw is insane. I got it up and running today. Yeah, I would agree. It's probably the biggest and best AI update that came out this year. And you can see we've now installed Grock as a search provider inside Open Claw. So to do that, all you do is you go to X, you grab an API key, you plug in the API key, and you ask openclaw to set up as a search provider. And then we should be able to use that instead of Brave Search. The good thing about using Grock as a search provider is like it it's constantly having updates. I tend to get most of my latest AI news from Grock as well. So, it's a really powerful way to use AI. What else we got here? GT5. 3 Codeex API is free to use even with a $20 subscription. Yeah, so you can you it's not the API, it's C O that you're using to login, right? That's how you use it. Alex Franco said, "Didn't the Chinese drop their version of open claw? Did you test out yet? " Which one are you referring to? I would be interested to know which one you mean. And then Antonio says, "I used Deep Seek V3 and I won and it's the cheapest model I found. You can actually get free models with open claw, so you don't need to pay for one. The other option that you have as well is like you can set up different personalities and sub agents with open router and as long as you got your API key plugged into open claw, you could use and route each of those APIs to a different provider depending on the task. So that's another option as well. But again, like you can use free APIs. I'm using free APIs today to do this. So that's basically it for using Open Claw. I've shown you how to add more personality inside open claw. the oneclick prompt that makes it a lot better. Would definitely recommend setting up. I've got a full guide inside

the AI profitable volume link in the comments description. If you go to the SAP update section here, we're updating this daily with new stuff. So, for example, we've got the full prompt you can copy and paste right here to make open claw 100x better. We've got the full section on agent teams and how to install AMP farms. We have the open claw update from February the 9th. So, every time there's a new update, um, even for example, like setting up Pony Alpha on Open Claw, every time there's a new update, we update this daily with everything that I've learned. And we usually include like video tutorials, step-by-step guides, etc. on exactly what you need to know about that update. Even for example, how to use OpenClaw notebook. So, feel free to get that link in the comments description. It comes with a 6-hour course on how to use OpenClaw 2. This is my AI automation community. And inside the air profit board number two, we show you three different ways to use OpenClaw for free. setting up the API keys for free. You can post inside the community, get any help and support whenever you need it. You can jump on the daily accountability groups if you want to stay focused and avoid shiny object syndrome. Inside the classroom here, we have all of our best trainings. So, you can see, for example, how I automate X and my AI avatar videos inside this section. If you want to learn how to go from beginner to expert with AI in just 5 weeks, you can learn that here, including how I build AI agents in under 5 minutes. Additionally, you can learn how to get more clients with the agency course. You can watch about the coaching calls that we do four times a week here. And you can learn how to rank number one with AICO. Plus, learn how to grow your own YouTube channel based on the training inside the classroom. So, feel free to get that link in the comments description or go to apiprofitboarding. com. Nobody says how do you give access to your computer to open claw. So, when you install it and you start asking it to like for example open up the browser or check out your local files, it will ask for permissions and then you just grant it the permissions. It's pretty easy to set up. Um, Antonia says, "For getting those free APIs, you need to pay some subscription. " No, no. So, like Miniax, you get 7 days free API. Alarm plus Kimmy K2. 5, you get free API within the token limit. And then Pony Alpha you get for free as well for open router. So, all three of those are free. Flavin says, "Does Pony have vision mode? " It's a good question actually. Let me just double check on Open Router. I don't think it does as far as I know. You can see in terms of all of the apps it's been used in, Open Claw is the number one. So, one that has been using the most and uh Kim K2. 5 is the number one model for using with Open Claw. A lot of people are using Trinity Large Preview as well. Not seen that one before. When did that come out? January 27th. Damian says, "I'm using 3. 2 at the minute for Deep Seek. I'm liking it so far. Not as flaky as Gemini Flash. " Sarah says, "I have a question. I installed OpenClaw inside VPS and I want to use it. Tell me in doing something using Chrome, but the Chrome extension um work just in local. Any idea how to solve this? " Yes, you would set up the you would set it up just inside the VPS instead. I guess that's what I would do and connect it that way. Flavven says there's so little information about it. It seems strange for it not to have it. How can we Well, it's just come out. That's why it's a stealth model, right? It's a stealth models typically there. There's never much information about how to use them or how they work and that sort of thing. How can we test it? I've got a full guide inside the AR profitable boardroom. So, if you go inside the classroom here, SB updates, you can see a full setup guide on how to use alpha pony alpha with openclaw and you can see a full video tutorial on how to use it as well right here. What are your thoughts about using Gemini Flash for OpenClaw? Honestly, I wasn't really that interested in using it. I like Gemini Flash inside anti-gravity. I'm not a big fan of it inside like chat bots and stuff. Don says, "We've got a ton of manual admin, emails, data entry, CRM, invoices. Can Open Claw handle this sent end safely? I mean like if you're worried about safety, why are you using Open Claw, right? That's what I would ask. And then yeah, I I mean like if you're if your focus is safety, then I would probably wouldn't use Open Claw, right? I'd use something um more secure that's more protected. Now, Gemini 3 Pro is good. Three flash gives slow progress and under that crash your system. Okay, good to know. And then Flavven says, "Antiggraity has 4. 6 now natively. " Yeah, I've done a video on that. Can you use it with o in open core? I've tend to what I tend to find is like when you're using um yeah so when you're using uh anti-gravity off it tends to break pretty quickly. I had to do a complete uninstall last time I tried that. What are the best lightweight models for small agents? So I mean you can like you could use Gemma 34B. That's probably the easiest way. Can you try codeex 5. 3? Yeahve I've tried it before. I did the o off with it. It was pretty good. How can it read videos to learn from your competitors? So, what you can do is set up Google. It's called a Google console, right? You get an API from Google console and then you can get it to scan your competitors that way as well. And then Damian says, I'm using it to set up and run a YouTube channel and makes videos using the Remotion API and does pretty much everything. It took a while to get going, but it's great. Now, is Quinn 3 Max thinking good via Open Router? I don't really like Quinn models that much. I like testing them, but I just don't use them dayto-day. I tend to find them quite slow. New command/collab experimental certain. Oh, not tried that.