The Top 10 Claude Code Plugins to 10x Your Next Project (June '26)

CChase AI
컴퓨터/소프트웨어창업/스타트업AI/미래기술

Transcript

00:00:00Every day, you're being told there's a new Cloud Code plugin that is going to change
00:00:03everything. And it's gotten to the point where there's so much noise out there
00:00:07that most of us have stopped listening entirely. But luckily for you, I've found some diamonds in
00:00:11the rough, 10 of them, in fact, and I am going to be sharing them all with you today.
00:00:16These are 10 Cloud Code plugin skills and CLIs that will actually make a difference on your
00:00:21next project. And no, none of them are the big names you already know about, like superpowers
00:00:25or front-end design. Instead, these are all relatively new. So unless you are as obsessed
00:00:30with this stuff like I am, I promise you will find something you haven't seen before.
00:00:34So the first one on our list is Graphify. And I actually did a full video on this yesterday.
00:00:39I will link that above. Now, Graphify is all about being able to point Cloud Code at some
00:00:45sort of repo, some sort of code base, and create a knowledge graph out of it. When I talk about
00:00:49a knowledge graph, I'm talking about something like this. This looks super cool, right? Visually
00:00:55stunning. But what it really is for Cloud Code is a map. It is a map showing Cloud Code how
00:01:00everything is connected and why. By creating this sort of knowledge graph with Graphify,
00:01:05we are able to have Cloud Code answer questions about a code base more effectively than without
00:01:10the graph. Normally it uses something called grep, but if we have the knowledge graph, we're
00:01:15able to answer the same questions with less tokens being consumed. Now, Graphify comes with
00:01:21a ton of commands. But luckily, when you install this thing, it includes a Graphify skill, which
00:01:26teaches Cloud Code what command it needs to use for what job. And there are some interesting
00:01:30ones in here. If you just do forward slash Graphify, that's going to have it run through
00:01:34the code base and create that knowledge graph you saw. But we can do things like add an Obsidian
00:01:38flag, and it will actually generate a brand new Obsidian vault for that repo you've just created.
00:01:44So if you're someone who loves Obsidian, but you want to add a little more power to the memory layer,
00:01:48Graphify makes total sense. Now, for those of you who are looking at this knowledge graph,
00:01:51like, hey, it looks neat. But what happens when I update the repo? It's a living, breathing thing.
00:01:56Well, Graphify thought about that. You can run this hook command, which will auto rebuild after
00:02:00every commit. So it stays updated. And auto rebuild sounds scary, like it's going to cost a bunch
00:02:05of APIs. That's not the case. It does AST only, which basically means the LLM isn't even
00:02:10involved. It's a deterministic thing. Now, the second skill on the list is actually a couple
00:02:15skills, and it comes from Matt Pocock, and that is Grill Me and Grill With Docs. These are two
00:02:22skills that are essentially plan mode on steroids. I love plan mode. You should be using it before every
00:02:27single feature or project. But Grill Me and Grill With Docs, I would argue, are a little bit better
00:02:31because it does what plan mode does. It asks you a ton of questions, so you and Claudco are on the same
00:02:36page, but it goes a bit deeper than the standard plan mode. And on top of that, these skills are
00:02:40pretty lightweight, so you aren't adding some huge token cost every time you spin it up.
00:02:46Now, the big sell for Grill Me and Grill With Docs and why these skills exist is we are trying to
00:02:50solve the problem of the agent not doing what we want. As Matt writes here, the most common failure
00:02:57in software development is misalignment. You think the dev knows what you want. You then see what
00:03:01they've built, in this case, Cloud Code, and you realize it didn't understand you at all. And if you've
00:03:06just been using plan mode and you expect you and Cloud Code to be exactly aligned on the same page after like
00:03:11three questions, you're going to be disappointed. And you've probably been disappointed in the past. And this is what
00:03:16Grill Me and Grill With Docs solves for. And piggybacking off that is skill number three, and that's Grill Me Codex.
00:03:22This is actually one I created. I did a video on this as well. And this takes Matt Pocock's Grill Me and Grill With Docs
00:03:29and adds a second layer to it. Because personally, I don't think it's enough for us to just get on the
00:03:36same page with Cloud Code in terms of our plan. I want to have a second set of eyes actually take a look
00:03:42at the plan Cloud Code has come up with. Yeah, you can ask me all the questions you want, but do you
00:03:47actually understand what Cloud Code built? Is it actually the best path forward? If you aren't an expert
00:03:53software engineer, you actually can't answer that question. Neither can Cloud Code, because
00:03:56Cloud Code cannot be trusted all the time to necessarily evaluate its own code correctly.
00:04:03This is something Anthropic has even brought up. So what this skill does is it brings in Codex to review
00:04:10the plan in a read-only sandbox. So you and Cloud Code have a back and forth, it creates the plan,
00:04:15then Cloud Code comes and takes a look. But it not only takes a look once, it has multiple rounds,
00:04:21up to five rounds, where Cloud Code and Codex go back and forth. And at the end, it's going to give
00:04:27you a thumbs up when they're both aligned. So we kind of get the best of both worlds. We get the best
00:04:31of Matt Pocock's grill me, and then we add a sort of Codex adversarial review. So you can be very
00:04:38confident that what you've come up with is the best path forward. And speaking of Codex, that brings us
00:04:42into plugin number four, which is the official Codex plugin for Cloud Code. What I just described
00:04:47in the last one, where we have this adversarial review and this back and forth, might be a bit
00:04:52much for some people. We don't need all that. We just want Codex to be able to come in and just do
00:04:56one pass at what we're looking at, or attack one feature. And the Codex plugin is perfect for that.
00:05:01This is from OpenAI themselves. So this isn't just some random guy putting stuff together. This is
00:05:07from, you know, the actual creators of Codex. And when you install this plugin, you're able to use it
00:05:12inside of the Cloud Code terminal. And it can do things like review, just do a basic review of what
00:05:17you've done. It has an explicit adversarial review, which goes deeper into some specific domains that you
00:05:23might not have thought of. And you can do things like Codex Rescue. So if you were building some
00:05:28specific feature and you wanted Codex just to work on that on its own, almost on the side, you can do
00:05:33that as well. And you can do this even if you aren't paying for OpenAI. There is a free tier for
00:05:39Codex now. Usage, you know, will be sort of an issue, but this is sort of like a great stepping stone
00:05:45if you don't want to go all in on Codex yet, but you want a second pair of eyes, but you don't want to
00:05:49do the whole free local thing. So highly suggest checking this out. It's been out for a couple months
00:05:54now, but to be honest, it's hard to find people who are really using this in their day-to-day
00:05:58workflows. Now, tool number five is Claude Obsidian. We all know about Obsidian, right? It's a free tool
00:06:04that allows us to organize our markdown files. And it really came into vogue in the AI space when
00:06:08Karpathy came out and talked about his wiki sort of style organization. Well, what if we had a plugin
00:06:15that organized it for us? That's exactly what this thing does. So this repo is meant to do everything
00:06:21Karpathy talks about when it comes to setting up your Obsidian Vault, but to automate the process. As it says
00:06:27here, you drop in the sources, whatever those are, Claude reads them, extracts entities and concepts,
00:06:32updates cross-references, and files everything into a structured Obsidian Vault. And the wiki or the
00:06:39vault itself continues to get richer and richer with each ingestion, which means we can more easily or we
00:06:46can more effectively have Claude code. Take a look at a large corpus of documents, whether they're
00:06:51structured or unstructured and give us correct answers. And what's also cool about this thing is
00:06:55the end of every session, Claude updates a hot cache. The next session starts with a full recent context,
00:07:01no recap needed. So if you're someone who's really interested in Claude and Obsidian and want sort of
00:07:07like an easy way to do all that for you and have all the organization be built in, highly suggest you
00:07:12check out this repo. Again, it is the Claude Obsidian repo. And speaking of Karpathy, we have to mention this
00:07:19one. This repo has 170k stars. It's a little bit older, but it's still one I wanted to bring up. And this is
00:07:25Karpathy's Claude.md file. And this is a great example of where we can make really effective changes with how we use
00:07:32Claude code in a rather simple manner. Because this is all it is. It's just a Claude.md file you can use. And it just talks about a few
00:07:40conventions that Claude should always follow. And they seem obvious on the surface, but there's also a reason
00:07:46this is so popular. The first one being think before coding. Number two, simplicity first, three surgical
00:07:53changes, four goal-driven execution. And that's pretty much it. Again, it seems obvious and you almost want
00:08:00to laugh like, okay, yeah, duh, you should do this. But there's something to be said about codifying this
00:08:04in a Claude.md file. That way, this is always followed for everything you do. And instead of
00:08:11having Claude try to go as fast as possible, these guidelines bias towards caution. Then moving on to
00:08:16tool number seven, we have Impeccable, which is my favorite front-end design tool in the game today.
00:08:22It's just one skill, but it encompasses 23 commands. And it's all about defeating the AI slop monster.
00:08:29Now, the amount of commands it has at its disposal are quite large, like all these different things,
00:08:35colorize, animate, onboard, distill, quieter. How are you supposed to know what these things are
00:08:40supposed to do? Well, obviously Claude code knows, but it helps for us to have an understanding of
00:08:45what's actually going on under the hood. And luckily, Impeccable makes it really easy because they have an
00:08:49entire website that breaks down with visuals what this actually brings to the table. For example,
00:08:56if I was wondering what Boulder does, and I will move this over here so it's easier to see,
00:09:00I can see what it looks like normally inside of Claude code, and I can see what it looks like with
00:09:05Impeccable. If I was looking at animate, again, we have Claude code design, and then we have
00:09:12Impeccable's design. We have this for 23 different sets, which is really, really cool. Another cool
00:09:17feature is that you can iterate on your UI with the live mode. So it'll start up a dev server. It'll
00:09:23actually bring up your website, and you can actually point at different elements of your
00:09:27website and edit them with Impeccable in somewhat of a live manner, which is really great because I
00:09:32know for me, I like to see the visual changes myself on the screen instead of always working
00:09:37in code. The number eight on the list is the Higgsfield CLI and Higgsfield MCP. I love this
00:09:41thing because I have a one-stop shop for every single AI image generator, AI video generator. And this is
00:09:48something I use multiple times a week. I do a lot of content, obviously, and a big form of content for
00:09:54me is things like carousels. And I'm able to create a full automation workflow, which brings in certain
00:10:02AI image generators with the Higgsfield CLI and just streamlines the whole process. Beyond that,
00:10:07things like Higgsfield MCP are also really great if you're doing front-end design work and want to
00:10:11bring an image or videos into your UI. Now, number nine on the list is Notebook LMPy, and I will
00:10:18never stop putting this in every single Cloud Code skill plugin CLI list I ever create because this is
00:10:23my favorite one. I use this literally every single day. This CLI lets you hook up Cloud Code to Notebook
00:10:29LM, and you get everything inside of Notebook LM and more with this tool. Why is it so good? Well,
00:10:35because Notebook LM is actually amazing, and it's free. You can offload tons of tasks to Notebook LM and
00:10:42Google servers, actual AI tasks that would otherwise cost you usage and tokens. And instead, Notebook LM
00:10:48does all of it for you. And like I said, the CLI actually gives you stuff that you can't do inside of
00:10:52the Notebook LM interface. Things like batch downloads, slide revisions, slide decks, PowerPoint,
00:10:57on and on and on. So I'm getting Notebook LM and more, and because it's a CLI, I can now integrate
00:11:03it into a number of skills. So I do tons of research related to YouTube that bring in Notebook LM
00:11:10Pi because Notebook LM itself is really good at processing and dealing with videos, especially
00:11:16YouTube stuff, because it's all under the Google umbrella. And last but not least, we have the official
00:11:20N8N MCP server. N8N still has a place in your AI toolkit. It isn't the dominant no-code platform it
00:11:27was like a year ago. But if you're someone, especially someone who works with clients and
00:11:32needs to create rather simple automations or workflows that the client themselves need to get
00:11:37hands-on with, there's no better way to do that than N8N. And the N8N MCP is the best thing on the
00:11:43market. We had really hacky options in the past, but with this tool, which you can use on the self-hosted
00:11:49version as well, by the way. So for all intents and purposes, it's free. You don't really ever have
00:11:55to even go into the N8N canvas except to like do a final check. This works really, really well. It's
00:12:01better than anything we've had in the past. So if you're someone who uses N8N in any capacity,
00:12:05you need to use this MCP. So those are the 10 Claude Code plugin skills and CLIs that I've been using
00:12:11in my projects lately. And I think they can definitely improve yours. So as always, let me know what you
00:12:18thought, make sure to check out Chase AI Plus down in the pinned comment if you want to get your hands
00:12:23on my Claude Code masterclass. And besides that, I'll see you around.

Key Takeaway

Integrating specialized plugins like Graphify, Grill Me Codex, and Notebook LMPy into Claude Code workflows reduces token costs and improves output accuracy through enhanced code mapping and adversarial plan review.

Highlights

  • Graphify generates a knowledge graph of a code base to reduce token consumption compared to standard grep methods.

  • Grill Me and Grill With Docs improve agent alignment by using deeper questioning patterns than standard planning modes.

  • Grill Me Codex adds a five-round adversarial review process between Claude and Codex in a read-only sandbox to validate project plans.

  • The Claude Obsidian plugin automates the extraction of entities and concepts from source files into a structured markdown vault.

  • Impeccable provides 23 specific UI design commands that allow for live, visual iteration on dev servers.

  • Notebook LMPy enables Claude to interface with Notebook LM to perform tasks like batch downloads and document processing while utilizing Google's free infrastructure.

Timeline

Codebase Mapping and Memory Layers

  • Graphify creates a knowledge graph of repositories to map code connections.
  • Deterministic AST-only rebuilding avoids LLM costs during updates.
  • Obsidian integration enables direct vault generation from code structures.

Graphify replaces traditional grep methods to provide Claude with a structured map of the codebase, which improves answer accuracy while lowering token usage. It includes an auto-rebuild hook that triggers after every commit using AST analysis, ensuring the map remains current without incurring additional API costs.

Agent Alignment and Adversarial Review

  • Grill Me and Grill With Docs deepen the initial planning phase to prevent project misalignment.
  • Grill Me Codex subjects generated plans to five rounds of adversarial review by OpenAI’s Codex.
  • The official OpenAI Codex plugin offers standalone review and rescue features via the terminal.

These tools address common software development failures where agents misunderstand user requirements. Grill Me adds depth to the planning stage, while Grill Me Codex forces an adversarial back-and-forth between two models to ensure the chosen technical path is optimal before execution begins.

Documentation and UI Design Optimization

  • Claude Obsidian automates the ingestion of documents into an organized, cross-referenced wiki.
  • The Claude.md convention file enforces four core principles: thinking before coding, simplicity, surgical changes, and goal-driven execution.
  • Impeccable offers 23 UI commands for live editing and visual feedback during front-end development.

Structuring knowledge and design is critical for maintaining project velocity. Automated vault updates in Claude Obsidian ensure context is preserved between sessions, while Impeccable streamlines front-end tasks by providing a browser-based interface to visualize and edit UI components directly through Claude.

Automation and Workflow Integration

  • Notebook LMPy offloads heavy document and video processing tasks to Google's free infrastructure.
  • Higgsfield CLI and MCP integrate AI image and video generation into standard development workflows.
  • The N8N MCP server enables self-hosted automation management without leaving the Claude environment.

Final workflow optimization involves leveraging high-utility external tools. Notebook LMPy provides a way to handle batch operations at zero token cost by utilizing Notebook LM, and the N8N MCP server creates a streamlined, no-code bridge for managing client-facing workflows.

Community Posts

No posts yet. Be the first to write about this video!

Write about this video