Insane GitHub Repos That 10x Your Codex And Claude Code Setup

0:00 Agents can build most of what you ask it for, but there's still a lot it can't do on its own. People have built tools that fix those gaps, and we went through them to find the ones that actually hold up. Eight of them changed how we work, and two of those have even been trending on GitHub. One of them lets you build those scrollable product landing pages that you've been seeing a lot. Another one solves the biggest problem with these agents, which is that they say a feature is finished when it actually isn't. If this is your first time, we're a software company, and this is our channel AI labs. And in this video, we're going over all eight GitHub repos, what each one changes in your setup, and how to use them. Now, you can take an object from an image and turn it into a 3D model made entirely from code, then add that model to your landing page, animate it in different ways, and let people interact with it directly on your site. The skill that does this is called image to 3js, and it's also one of the top trending GitHub repos. And because the entire model is made from code, your agent can adjust it the way you want. You can ask it to change the colors, adjust the lighting, or add without having to use extensive software to animate 3D models yourself. Now, to use it, you just need to give your agent the image you want it to recreate and ask it to use image to 3js to turn it into a 3D model. The agent will then build the object in stages. As it builds, it compares the model against your image. If it finds differences between the model and your image, it will fix them before adding more detail.

1:23 This whole workflow burns a lot of tokens and takes a long time. The model might not be as detailed as you want on the first try, but you can keep asking the agent to improve specific parts and get closer to the level of detail shown in the product demos. Another problem with agents is that they mark a task as complete even though it isn't. They usually check whether the app loads correctly and whether the design was implemented properly, but there are still things that can fail behind that.

1:47 The only way to catch those is to give the agent more access to what's actually happening in the browser. So, to solve that problem, there's a tool called Reticle. Once the agent says it's finished the app, this tool checks whether the app actually works the way it should. It lets the agent follow what happens inside the app as it runs, so it can find out why something isn't working. Once it's installed, whenever you ask your agent to build an app, it will use this tool to check what it built, even if you don't mention it in your prompt. The agent automatically opens the app in the browser and uses it the way you would, so it can check whether the app actually works. After the agent has tested the app through Reticule, it basically gives three conclusions on each part it tested.

2:26 These three conclusions are either it worked, it didn't work, or there wasn't enough information to tell. When something fails, Reticule explains what went wrong, so the agent can fix the problem and try again. And when you open the app yourself, you'll see a pop-up in the corner for Reticule that you can interact with. You can also look through the previous checks to see what went wrong and how the agent fixed those problems. But before we move to the next repo, it would be great if you subscribe to the channel and hit the hype button.

2:51 This small gesture of support goes a long way for us. You've probably noticed how long the replies from models have been getting lately, especially the summaries they give at the end of their work. This is more prominent in Claude models. Some of that explanation isn't even helpful because of how it's structured. The replies get so long and the words are laid out so oddly that it becomes hard to understand what the agent is trying to say. So for that, there's a tool called Chisel that tells the agent to use fewer words and avoid writing code the task doesn't need. The aim is to reduce the tokens the agent spends on unnecessary explanations and code. Chisel's author has published test results as well. In his example of changing how a search box works, the agent used about 1,500 tokens for its answer without Chisel. With Chisel, the same task used about 600. You might have already heard of tools like Caveman and Ponytail that do similar work. Chisel's author compared those tools as well, and the agent used fewer tokens overall with Chisel. It installs as a plugin for Claude code. Once you install it, you'll get four skills along with hooks. For those who don't know about hooks, they're commands that run automatically at particular points in a session.

3:56 Chisel installs three of them. One runs when the session starts, another when you send a prompt, and the last one after the agent uses a tool. But if you're using Claude code, you get an extra benefit, which is that Chisel also reduces the amount of information the agent has to read. When tools send information back, Chisel removes repeated or unnecessary text before the agent reads that information. This helps keep the context from filling up with text the agent doesn't need. And with less text to read, the agent uses fewer tokens, which also helps reduce the cost of the session. With these skills installed, whenever you give any task to the agent, it will return a much smaller summary than it used to, using more direct wording, so that you don't have to struggle to understand what it means.

4:37 The outputs are clearly structured, first telling you what it generally did, then the key things it changed, and finally what it skipped, all laid out properly. There are a lot of skills for improving your designs, and each one comes with its own design principles, but that's exactly what makes it hard to choose, because there are so many of them. And figuring out which one fits the style your design needs takes real work. For solving this problem, there's a tool called UI skills, which is a library of design skills from different authors. It lets the agent search through those skills and gets whichever skill fits the design you want the best. It's available as both an MCP and a tool you run from the terminal, and you can choose between the two. We went with the MCP option because its tools stay available in the agent session. With the terminal option, you need to tell the agent when to use UI skills, either in your prompt or in your project instructions. So, once you've connected the MCP, it'll be listed among your agent's MCP tools. After that, you can ask your agent to use UI skills for the design you want to make. The agent first searches the library for skills that fit what you asked for in the prompt. From there, it narrows down to the relevant ones and gets their instructions to use while building the site. So, you don't have to choose and install every design skill yourself before you start. But, before we move on to the next one, let's first have a word by our sponsor, Manufact. Getting your app to actually work inside Claude means shipping an MCP server, and that's a whole project on its own. Between a protocol that keeps changing, plus hosting and deployment, that's easy to get wrong. Manufact made the entire process faster and easier. We built the MCP app with their open-source MCP use SDK, connected our GitHub, and now every commit push deploys the server automatically. Then we set up a custom domain, pasted the URL straight into Claude, and our app's real UI showed up live in the chat. We ran it against ChatGPT 2, and so we could test both clients. And honestly, seeing our own app running inside Claude was so gratifying. Manufact also runs publishing checks, so your app is ready for the Claude Connector's Marketplace and the ChatGPT App Store, not stuck in review. It's like Versel, but for MCP.

6:41 And MCP use has over 10,500 GitHub stars and over 9 million downloads. Deploy your first MCP server for free at manufact.com. If you haven't explained exactly how you want your app to work, the agent will decide those details for you. Oroboros is a tool that helps the agent work through those details before building the app. It then checks the app and sends any problems back to the agent to fix. To install it, you need to follow the setup commands for the agent you're using. It includes a command that you run from the terminal, and the setup connects that command to your agent, so it can run the workflow for you. Once you've set it up, you need to describe the app you want to build and ask the agent to use Oroboros.

7:20 It starts by interviewing itself. It generates questions about your request, then answers them from what you've already described and what it can see in your project. It only fills in small missing details on its own when those choices are easy to change later and stay within what you asked for. If a decision could change what the app is supposed to do, it pauses and asks you. It also records any assumptions, so you can see what was decided. Those answers become a plan that describes what the finished app needs to do. Before building starts, Uura Borus reviews that plan and works on any problems it finds, and the agent starts building only after the plan passes. Once the app has been built, Uura Borus checks whether it does what the plan says. The agent's instructions for building the app never include how those checks are run or what results they expect. If something doesn't work as planned, Uura Borus tells the agent to fix that part. The agent keeps what already worked and those checks run again to make sure the changes haven't broken anything. You can also open the Uura Borus dashboard to follow the agent's work as it happens.

8:17 If the agent stops making progress or reaches its limit on attempts, Uura Borus will tell you why the work stopped. There are a lot of design skills available, but fewer are focused on how iPhone apps should look and behave. When you ask an agent for those, it tries to recreate the look of liquid glass, which is the glass-like style Apple uses for buttons and menus by adding blurs and backgrounds. But the problem is that you get something closer to a transparent button than the real thing, even though the agent tells you it's the same, and the apps it builds don't keep up with the functional changes Apple has been making either.

8:51 So, for iOS apps, there's a repo called Swift UI skills. It contains the instructions for building your screens properly with Swift UI, which is Apple's own system for building them, and it's kept updated with the latest changes Apple has been making. There are two skills in the repo. The liquid glass skill tells the agent how to use Apple's built-in buttons and menus in iOS 26 and which common mistakes to avoid. The other one covers how to arrange your app's content for the iPhone Duo, which is the new phone that Apple just launched recently, so your design works at that wider size as well. Once you've installed the skill you need, you can ask your agent to follow its instructions while building your app. It will then use Apple's actual patterns, which gets you much closer to a native app than the agent manages on its own.

9:34 When you've created a lot of skills, it gets hard to know which skills are still helping the agent do better work. This becomes harder after a model update because the newer model may already do what a skill tells the agent to do without needing those instructions. So, a skill that helped before might now be adding to the token cost without improving the work. You might think Claude Code's new skill doctor command already checks this. That command shows how many tokens each skill takes up in the context and which skills the agent hasn't used, but that doesn't tell you whether a skill actually helps the agent do better work. Caliper is a tool that basically tests if the skill that you've created actually does what it was made for and is actually having any impact on your workflow or not. It can also compare the agent's work with and without the skill to see whether those instructions actually help. Caliper runs from the terminal and the repo includes two skills that help your agent use this tool. The first one is grill skill, which interviews you about what your skill should do. So, the tests actually reflect what you intended. The second is the evaluate skill, which helps your agent create and run those tests, which are called evals. Each test includes a prompt and a description of the result expected from that prompt. The agent saves those tests alongside your skill and Caliper keeps records of the runs in a folder called {dot} Caliper, so you can look through the results later. So, to use Caliper, you just need to ask your agent to use it on a skill you've created. The agent will then invoke the evaluate skill first and prepare the tests for the skill. If it gets stuck in writing tests, it will then invoke the grill skill and interview regarding it.

11:05 Once the tests are written, it will launch multiple runs with different prompts, few using skill and few without the skill, and see how the impact compares. At the end, your agent gives you a report showing what happened across multiple runs and what you can do next. The report helps you see which tasks the skill handles reliably and where the skill still needs work. Comparing runs with and without the skill also shows whether the improvement is worth the extra tokens. Caliper can also make the same comparison for MCP servers and other instructions in your setup. So you can check more than just the skills you've installed. Anti-slop is basically a tool that's built to remove all the bad coding patterns agents tend to use when writing code. It looks for specific coding mistakes and reports them to the agent so they get corrected without you having to read the code yourself. It checks the code using a set of rules written by the author.

11:53 Some of them catch unnecessary work that can slow the app down while others reflect how he prefers to write code. You can follow the instructions in the repo to set up anti-slop in your project. So when you have the skill installed, whenever you give your agent any task, it will run these checks and fix all the issues that it comes across. You can also ask it to run a final review once you have completed what you're working on. Now everything we build and show you in all our videos can be found in AI Labs Pro, which is our community. So if you found value in what we do and want to support the channel, this is the best way to do it. The link's in the description. That brings us to the end of this video. If you'd like to support the channel and help us keep making videos like this, you can do so by using the Super Thanks button below. As always, thank you for watching and I'll see you in the next one.