{"id":"2091484130290888729","url":"https://x.com/charliejhills/status/2091484130290888729","text":"","author":{"name":"Charlie Hills","username":"charliejhills","avatarUrl":"https://pbs.twimg.com/profile_images/1930709528829423616/U7t4WBAU_200x200.jpg"},"createdAt":"Sun Aug 23 11:14:09 +0000 2026","engagement":{"replies":9,"retweets":8,"likes":89,"views":216324},"article":{"title":"Give me 10 minutes. I'll teach you 80% of Claude Code.","previewText":"On the 3rd of March 2026, I opened the terminal, typed \"claude\", and had no idea what I was doing. 173 days later, it runs (almost) 100% of my content business.\nI'm not a developer. I'm a marketer who","coverImageUrl":"https://pbs.twimg.com/media/HQZxUDebAAAMPPg.jpg","content":"On the 3rd of March 2026, I opened the terminal, typed \"claude\", and had no idea what I was doing. 173 days later, it runs (almost) 100% of my content business.\n\nI'm not a developer. I'm a marketer who discovered the power of Claude Code for content. I might sound strange to say that, because coding is for developers, right? Wrong.\n\nCoding is for everyone.\n\n![](https://pbs.twimg.com/media/HQZi3YMbAAACTf3.jpg)\n\nYou've heard the term \"vibe coding\" before, I'm sure. The practice of describing what you want to AI and it'll build whatever you want. The same applies to marketing.\n\nEvery day I'm \"vibe designing\" or \"vibe marketing\" with Claude.\n\nI've spent the last 173 days mastering the practice, so in this article I hope to offer you a glimpse of what it has done for me and what it could do for you, too.\n\nThis covers everything I have learned in 173 days.\n\nAnthropic just dropped three new features worth knowing about if you are running Claude Code properly.\n\nIf you’re already comfortable in Claude Code, start here.\n\nIf you’re still learning, read the article, then watch :)\n\n[3 New Essential Claude Code Updates (Nothing To Install)](https://www.youtube.com/watch?v=plckLUuVIik)\n\n## First, what this thing actually is\n\nClaude Code used to terrify me.\n\n1. I'm a marketer by trade.\n\n1. I'm not (actually) technical.\n\n1. I'm definitely not a developer.\n\nSo I thought, why should I be using Claude Code?\n\nWell, Claude Code has changed the way that I run my content business. And it's not just for building applications. You can use Claude Code to create content, too.\n\nIt reads, writes and edits your files. It calls APIs. It browses the web.\n\nIt spawns workers to run jobs in parallel. It runs scripts you never wrote.\n\nYou do not need to be a developer. The barrier around tools like this is largely gone, and a few hours of learning-by-doing gets you real results.\n\n## #1 Get it running (about two minutes)\n\nDay 1. I opened the terminal and typed “claude” with no idea what I was doing. Nothing I built in the first month compounded. Every session started from nothing and I re-explained myself.\n\nGet a paid Claude plan first. Pro at $20/month is enough to start. Max gives you more usage if you end up running heavy workflows.\n\nInstall it with one command in your terminal, from code.claude.com/docs.\n\nTakes about two minutes.\n\n![](https://pbs.twimg.com/media/HQZjcJraUAAdRbI.jpg)\n\nIf you've never opened it, you access Claude Code through the terminal or through the Claude desktop app. I run mine through the terminal.\n\nThe desktop app is pretty good. Start there if you’re new.\n\nType claude. It prompts you to log in.\n\nNow navigate to a folder you actually work in. You're running.\n\n![](https://pbs.twimg.com/media/HQZj6WPakAAoyr4.jpg)\n\nWhat you need to build out here is a CLAUDE.md file. This is basically your background on who you are, what you like, and how you exactly want Claude Code to work.\n\nOne of the best practices here is to simply type this:\n\nThis will initialise a new CLAUDE.md file for you.\n\nBut you don't have to follow this step. To be honest, you can just talk naturally to Claude Code, and then ask it to update your MD file in retrospect, based on the learnings that you make along the way.\n\nThat second way is the one I'd actually recommend, because you don't yet know what's worth writing down.\n\n## #1.5 Then open /config\n\nDon’t judge me, it’s day 173, and I only discovered this panel today.\n\n/config is where Claude Code keeps its settings. About forty of them.\n\n![](https://pbs.twimg.com/media/HQZkQJ2aoAAV2ZN.jpg)\n\nI'm not going to pretend I've tuned them. Most are still on whatever they came with. The point is that on day 173, writing a guide about this thing, there was a whole panel I'd never opened.\n\nThat's the thing about Claude Code. You find something new and work out how to use it every single day. With Claude Chat and Cowork you're fairly limited in what you can actually do. The ceiling is (almost) uncapped.\n\n## #2 Plan before it builds\n\nYou can slash command /plan, and Claude Code will build out a plan before it builds anything. I'd highly recommend using that command if you're totally new to Claude Code, like I was in March.\n\nThere are two ways in. Type /plan, or hit shift + tab. Same mode, and shift + tab cycles straight back out again.\n\nPlan mode is read-only, and it is far cheaper than rebuilding. A failed build costs more than ten minutes of planning ever will.\n\n![](https://pbs.twimg.com/media/HQZkWY2bkAAwlXR.jpg)\n\nThis is the alignment step that runs before Claude builds anything. Claude proposes a layout, asks me what I want changed, and then it codes.\n\n![](https://pbs.twimg.com/media/HQZkcxLaQAAHDOd.jpg)\n\nPro tip: if you're mid-task and the output starts feeling off, drop back into plan mode to realign on where you actually want to take it. This is way better than letting Claude run down the wrong path and burning tokens.\n\n## #2.5 Auto mode\n\nWhen I first started using Claude Code you had to bypass permissions, because otherwise it asks you a tonne of questions and it is genuinely annoying.\n\n![](https://pbs.twimg.com/media/HQZlDaEbcAAdOAA.jpg)\n\nThat is probably the first thing that blocks you when you get started. Not the terminal or the setup. Just being interrupted every thirty seconds.\n\nThat is what auto mode is for, and it is now the default. It is safe to use, and honestly, as long as you are not doing anything too sensitive, you are good.\n\n![](https://pbs.twimg.com/media/HQZlJmEbkAEqfDm.jpg)\n\nThe reason I am comfortable saying that is what sits underneath it. In auto mode every thing gets checked before it runs, and it deliberately ignores your own allow-list rather than trusting it. Shift + tab cycles you between manual, auto mode, and plan mode.\n\n## #3 Install things\n\nThis is the section I’d move to day one if I could start again.\n\nI spent my first month building my own versions of things that already existed, were better, and were free to install.\n\nType this:\n\nThat opens the marketplace.\n\n![](https://pbs.twimg.com/media/HQZlYGzbYAAKoga.jpg)\n\n## 1. Superpowers\n\nTo be honest, I only use this for brainstorming, and I do not even use that much.\n\n## 2. gstack\n\nI use this for one thing. The agent browser.\n\nIt opens a real browser and reads the live web, so you get what is on the page today.\n\nYou find it on GitHub, click code on the top right, and copy the installation command.\n\nMy favourite thing about this workflow is that I built a system that writes my newsletter with me. Claude Code can now do a version of this itself (without gstack).\n\nIt opens tabs in your own Chrome, the one already on your desktop, and it carries your logins with it. So it can work inside Gmail, Google Docs, Notion, anything you are signed into, with no connector set up at all.\n\n## 3. HyperFrames\n\nWrites HTML and renders it as video. I use it for the HTML video editing.\n\nAlmost every YouTube video I ship comes out of this. I describe the overlay, Claude writes the HTML, it comes back as an mp4.\n\n## 4. Claude HUD\n\nThis one sits at the bottom of my screen every time, so I can see my context and my usage.\n\nWhat it’s doing, and what it’s spent.\n\n![](https://pbs.twimg.com/media/HQZmGUjacAA6C1A.jpg)\n\n## 5. Caveman\n\nI installed this on day 72 to get short replies.\n\nIt gives no preamble or summary, just the answer.\n\n![](https://pbs.twimg.com/media/HQZmdlla4AA_gbc.jpg)\n\nPick the one for your job\n\nThose five are mine because of the job I do. Yours will be different.\n\nIf you want somewhere to start that is not my list, Matt Pocock has a skills repo that is going up fast. 224k stars: [github.com/mattpocock/skills](http://github.com/mattpocock/skills)\n\nWhich is the exact problem I got tired of. Every week someone asks me which repo to install, and whether you install it depends on what they do all day.\n\nSo I built the Claude Code GitHub Repo Bible.\n\nThere are 268 repos worth installing, in seven chapters, every link checked live and ranked by stars.\n\n## #4 Connect tools\n\nAll your connectors carry over from claude.ai. You don’t set them up again.\n\nAnd I only keep the tools I genuinely use every day.\n\nConnectors are one of the most useful features in Claude.\nConnect Gmail and your calendar, and Claude can read both.\n\nAsk it for a summary of your week and it actually knows what is in there. Then it can act inside those apps on your behalf.\n\nTo add one, go to customise, then connectors. There is a list of pre-built ones to browse.\n\nConfigure them in the web interface, and your Claude Code sessions inherit them.\n\n![](https://pbs.twimg.com/media/HQZoVUPbQAARbha.jpg)\n\nOr you can add custom connectors, like I’ve done.\n\nGoogle the tool name plus \"MCP\". Most things have a connector by now.\n\n![](https://pbs.twimg.com/media/HQZorFIacAA8_ZD.jpg)\n\nIf yours doesn’t, use Zapier. It carries 9,000+ integrations, so whatever you run is in there somewhere.\n\nThere’s really no excuse anymore.\n\nWorth two minutes on day one, because you are probably already holding more than you think:\n\nI don’t leave the terminal, every tool gets called from there.\n\n## #5 Brief it\n\nHonestly, I didn’t really start with a CLAUDE.md.\n\nI started with Boris Cherny’s, because it is the best starting point there is. He built Claude Code. I covered that here: [Claude Code (for normal people)](https://charliehills.substack.com/p/claude-code-for-normal-people).\n\nCLAUDE.md is the single most important file in your entire setup.\n\nEvery project folder, skill and workflow you build inherits one of these.\n\nIt’s markdown, which is just text. It’s basically a system prompt. That’s all it is. It’s nothing complex. Mine holds my brand rules, my voice and my output format in one place.\n\nWrite your CLAUDE.md before you write a single prompt, and everything else adds on top of it. Where I went wrong was what I thought it was for. I treated it as the place everything lives. All the rules, all the context, every correction.\n\nIt’s your front desk, not your filing cabinet.\n\nThen you point Claude at a specific folder. This is separate from your global one.\n\nGlobal CLAUDE.md is the instructions that apply everywhere.\nLocal CLAUDE.md is the instructions for one project.\n\nIt’s basically the same as your custom instructions in ChatGPT or Claude, and your project instructions in the chat interface. Same idea, just sitting in a file instead of a settings panel.\n\nContext is what keeps the two from bleeding into one another.\n\n![](https://pbs.twimg.com/media/HQZpIbubkAAmFcZ.jpg)\n\nKeep them apart and neither one bloats.\n\nMix them and your global file becomes a landfill. That is exactly what mine became, because every correction feels global at the moment you make it.\n\nAlmost none of them are.\n\n## #6 Folders\n\nI have one folder per social platform. That is my system.\n\nLinkedIn got its folder on day 24. YouTube on day 65. Instagram and Substack on day 93. It took me three months to cover four platforms, one at a time, as each one got busy enough to need its own rules.\n\nThe restraint is the useful half. Do not open the second folder until the first one is genuinely busy enough to need its own rules.\n\n![](https://pbs.twimg.com/media/HQZpO49aoAEAOIz.jpg)\n\nI have been slowly rolling out folders as I build out my content ecosystem.\n\nYou don’t need to set up folders on day one. Perfect the most important thing first, which in my case was LinkedIn. Then gradually roll out folders, skills and MD files to the rest of the system.\n\nLinkedIn holds 1,586 files and 1,037 lines of rules. X is three files.\n\nA file that only loads for one job can be specific in a way a global file never can.\n\nNever have one giant file.\n\nPaste this, inside a project folder:\n\n![](https://pbs.twimg.com/media/HQZpd95b0AAlxoG.jpg)\n\nI (finally) wrote my Instagram project’s CLAUDE.md writing this newsletter :)\n\n## #7 Memory\n\nIf you don’t have memory enabled, this is the simplest fix you can make to your Claude experience.\n\nMemory files I wrote per month: March 11, April 48, May 82, June 129. Then July, 590. You see the sharp rise in my memory. It begins the month after I started writing sessions down properly.\n\nJust tell Claude to update its memory and it will.\n\nIt also does a chunk of this on its own now. Auto memory means Claude writes down what it learns as it works, across sessions, without you asking. Just type /memory to enable it.\n\n![](https://pbs.twimg.com/media/HQZpv9AbkAAfKIh.jpg)\n\nHere’s example of it in action:\n\n![](https://pbs.twimg.com/media/HQZp7naawAALBZw.jpg)\n\nThis makes the folder job smaller than it was when I started.\n\nYou are steering what it keeps, not typing it all out yourself.\n\nYou don’t need a complex memory system. You don’t need to learn graph engineering. You just need folders, made the way you want them to work.\n\nSay you are doing design work. You need your references and your brand guidelines. You just need to subcategorise them.\n\nPaste this:\n\nThen add this one, so it keeps itself up to date:\n\nThe more Claude knows about you, the less you have to re-explain yourself.\n\nClaude will make mistakes. It takes time to ramp up, but you’re building a system.\n\n## #8 Skills\n\nAnything you've retyped twice is already a skill.\n\nNot the cleverest thing you do. The most repeated one.\n\nSkills are how I turn the repetitive work in my business into one-line commands I fire from any chat. They're how I take a 200-word prompt and turn it into an eight-character command.\n\nThe only way I make new ones is this. Chat with Claude until the workflow is dialled in, then ask it to save the thing as a slash command.\n\nSo paste this, right after you've had a Claude that worked:\n\nChoose the trigger word deliberately. Too common and it fires in the middle of unrelated work, which is genuinely worse than it never firing at all.\n\nI can’t remember the first skill I made. It is just something you intuitively do as you are doing a task, and then you ask Claude to make it into a skill. If you spend enough time and perfect a process with AI, make it a skill. If you do it twice, make it a skill.\n\nThere are three ways to get one, and none of them involve writing a file yourself.\n\n1. Get to the end of something that worked and say, save this as a skill.\n\n1. Or give Claude a lot of context and say, I need a skill that does this.\n\n1. Or record one. I walked through that here: [Record a skill in Claude](https://charliehills.substack.com/p/record-a-skill-in-claude).\n\n## #8.5 The ones I (actually) type\n\nA command ships with Claude Code. A skill is a file you made. Both start with a slash, which is the entire reason people mix them up.\n\nI went back through every prompt I have typed since January. Twenty-eight thousand of them, across two hundred and seventeen days. It is not the list I would have given you from memory.\n\nThree built-in ones I genuinely reach for:\n\n/plan is the only thing that survived from day one. I typed it at 16:57 on my first afternoon, and again this morning.\n\n/goal I use most widely, a hundred and eight times across thirty-eight days. It goes at the end of a prompt, not the start:\n\nOne more, for anything I need properly researched:\n\nIt sends five searches out at once, then argues with itself, and anything it can refute twice gets cut before you see it. What comes back is a short report where everything left standing survived a fight, with the sources next to it.\n\nThen the three I built myself, which I use every working day:\n\nSave progress writes what the session learned into my files.\n\n![](https://pbs.twimg.com/media/HQZrkX-bQAEf3Uh.jpg)\n\nClose finds every loose end and either finishes it or hands it back as one decision.\n\n![](https://pbs.twimg.com/media/HQZrq5aa0AA-cuy.jpg)\n\nShow me puts whatever I have just made in front of me in the browser, instead of describing it.\n\n![](https://pbs.twimg.com/media/HQZr8uKaIAAuTAv.jpg)\n\nI’ve used /save-progress 103 times. /close, 80. /show-me 131 in the 11 days since I built it. More than any command that ships out the box.\n\nThat is the argument for building your own. The ones you make for your own friction get used harder than anything that arrives in the product.\n\nAll three are free and public: [save-progress](https://github.com/charlie947/save-progress), [close](https://github.com/charlie947/close) and [show-me](https://github.com/charlie947/show-me).\n\nAnd the one I cannot give you:\n\nIt runs my LinkedIn graphics from idea to finished render. The only thing I have built and will not publish, because it is becoming a product.\n\nIf you want it when it goes live, sign-up here: [https://growth.charliehills.io/are-you-a-fit](https://growth.charliehills.io/are-you-a-fit)\n\n## And keep them small\n\nDon’t have a monster skill. If a skill is really comprehensive, Claude is not going to read all of it. What you want is a CLAUDE.md that directs to a skill, and that skill directs to the next one, and it cascades. So every skill is a specialist, not an encyclopedia.\n\nI have 214 on this machine. The average one is 539 lines, and the biggest is 3,153. That one loads its whole body every time it fires, whether I needed the last 2,000 lines or not.\n\nPaste this:\n\n> Two things, and do not change anything yet.\n\n1. List any skill of mine doing more than one job, and tell me exactly where the split is.\n\n2. Look back over my history and tell me which of my skills I have never once invoked. Rank that list by how long ago I made them.\n\n![](https://pbs.twimg.com/media/HQZsl72bgAAAdlS.jpg)\n\nThe other half of that is the ones you never fire.\n\nOf the 214 sitting on this machine, only 72 have ever actually fired. The other 142 have never run once.\n\nI pruned mine properly a fortnight ago and wrote up exactly how in [Delete your CLAUDE.md](https://charliehills.substack.com/p/delete-your-claudemd). Same job, same method, whether it is instructions or skills.\n\n## #9 Make it fire without you\n\nRule of thumb: CLAUDE.md guides. A hook forces.\n\nI wrote six hooks in my first 150 days and nine in the last 23. I spent five months writing requests and wondering why they only worked most of the time.\n\nEverything in your CLAUDE.md is a request. Claude reads it, weighs it against everything else you wrote, and usually complies.\n\nUsually.\n\nA hook is a script that runs automatically at a fixed moment, and Claude doesn't get a vote. Straight from the docs: to block an action regardless of what Claude decides, use a PreToolUse hook.\n\nThe test: would you be annoyed if this got ignored once?\n\nIf yes, it shouldn't be a sentence in a file.\n\nThe word is the hook. What it goes and gets is the skill.\n\nThe simplest one you could build:\n\n> When I say the word “LinkedIn”, write me a LinkedIn post.\n\nThat’s a hook.\n\n![](https://pbs.twimg.com/media/HQZswWaa8AACQhq.jpg)\n\nI never tell it to. I just say the word/s.\n\nThis is where hooks and skills meet. The word is the hook. What it goes and gets is the skill. You cannot see that pairing anywhere in Claude, which is why it took me so long to work out I already had loads of them.\n\nStart by finding out what you already repeat. Paste this:\n\n> Help me work out what I should turn into a hook.\n\nLook back over our recent conversations and find:\n\n1. Things I ask for over and over, in roughly the same words.\n2. Times I had to tell you something you should already have known.\n3. Times you did something I had to correct more than once.\n\nFor each one, tell me:\n- the word or phrase I actually say when I want it\n- what you should go and read, or do, the moment I say it\n- whether that word is unique enough that it won't fire by accident\n\nGive me your top three, ranked, with the best one first. Do not build anything yet.\n\nIt read 15,109 prompts I had typed over 45 days. My most repeated phrase was “show me”, 424 times, and nothing was firing on it.\n\n![](https://pbs.twimg.com/media/HQZtyBDaIAATO7O.jpg)\n\nThen just tell Claude to turn the top one into a hook (no special command or config needed). You describe it, it builds it, and it shows you before it saves anything.\n\nWorth adding this to your CLAUDE.md while you’re there:\n\nIf you want to read about it in more detail, you can visit:\n\n https://charliehills.substack.com/p/give-me-10-minutes-ill-teach-you\n\nStay curious, stay human and keep creating\n\n— Charlie"},"adhxContext":{"savedByCount":1,"publicTags":[],"previewUrl":"https://adhx.com/charliejhills/status/2091484130290888729"}}