Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

THIS GUY AUDITED 926 CLAUDE CODE SESSIONS AND FOUND MOST OF THE TOKEN WASTE WAS ON HIS SIDE everyone is blaming anthropic for the limits, so he decided to actually look at the data 858 sessions, 18,903 turns, and $1,619 estimated spend across 33 days here's what he found:... 1\ one default setting was burning 14,000 tokens per turn Claude Code loads the full JSON schema for every tool into context at session start. whether you use them or not. 20,000 tokens of tool definitions sitting there on every single turn. the fix: one line in your settings.json "ENABLE_TOOL_SEARCH": "true" context dropped from 45K to 20K instantly. across 858 sessions that one setting was wasting an estimated 264 million tokens 2\ cache expiry is the single biggest waste 54% of his turns came after a 5+ minute idle gap. every one of those turns re-processed the entire conversation at full price which caused a 10x cost jump you go grab coffee. come back 5 minutes later. type your next message. everything rebuilds from scratch. the context didn't change. you didn't change. the cache just expired. 12.3 million tokens wasted on idle gaps alone 3\ 42 skills loaded. 19 of them used twice or less across 858 sessions. every one of those skill schemas sat in context on every turn eating tokens for nothing. 4\ 1,122 redundant file reads where the same file was read 3+ times one session read the same file 33 times. he ALSO built a full token auditor dashboard that shows you exactly where your waste is coming from 19 charts, opens in your browser, free AND open sourceshow more

Om Patel

26,260 subscribers

298,746 views • 4 months ago •via X (Twitter)

Education Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

someone just rebuilt claude code in pure python and made it ~230x cheaper to run it's called clawcodex. a full from-scratch port of the claude code agent loop. 230k+ lines of python, MIT licensed it keeps your request prefix byte-stable, so DeepSeek's prompt cache covers your entire system + tools + history on every turn cache-hit input then bills at ~$0.0435 per 1M tokens the same 1M tokens on Claude Fable 5 runs you $10 so the longer you code, the more the cache pays off what you actually get: → 25 LLM providers (Claude, GPT, DeepSeek, GLM, local ollama) → 1M token context window → real agent loop: tools, skills, REPL, session history → one-line install, mac/linux/wsl and it's sitting at 677 stars. basically nobody is looking yet bookmark this before it gone

someone just rebuilt claude code in pure python and made it ~230x cheaper to run it's called clawcodex. a full from-scratch port of the claude code agent loop. 230k+ lines of python, MIT licensed it keeps your request prefix byte-stable, so DeepSeek's prompt cache covers your entire system + tools + history on every turn cache-hit input then bills at ~$0.0435 per 1M tokens the same 1M tokens on Claude Fable 5 runs you $10 so the longer you code, the more the cache pays off what you actually get: → 25 LLM providers (Claude, GPT, DeepSeek, GLM, local ollama) → 1M token context window → real agent loop: tools, skills, REPL, session history → one-line install, mac/linux/wsl and it's sitting at 677 stars. basically nobody is looking yet bookmark this before it gone

Argona

51,369 views • 1 month ago

A 22-YEAR-OLD FROM BEIJING BUILT A SECOND BRAIN WITH CLAUDE, MAKES $14,000/MONTH AND HIS AI REMEMBERS EVERY DECISION HE'S EVER MADE Most people open ChatGPT, ask a question and start from zero next session, re-explaining the same context every single time and losing 15–40 minutes per session doing it. He built infrastructure where Claude reads, writes and connects everything in Obsidian automatically so it never resets and compounds with every conversation. This morning he needed a client strategy from 3 months ago. Claude pulled it up instantly and drafted a full email reply in under 20 minutes using context it had been building for months across 6,000+ sessions. Every session adds to the last one and the system now saves him 2–3 hours every single day - 30–90 minutes from automated morning briefing alone, before he even opens his laptop. repo: The repo is free. The advantage it creates is not.

A 22-YEAR-OLD FROM BEIJING BUILT A SECOND BRAIN WITH CLAUDE, MAKES $14,000/MONTH AND HIS AI REMEMBERS EVERY DECISION HE'S EVER MADE Most people open ChatGPT, ask a question and start from zero next session, re-explaining the same context every single time and losing 15–40 minutes per session doing it. He built infrastructure where Claude reads, writes and connects everything in Obsidian automatically so it never resets and compounds with every conversation. This morning he needed a client strategy from 3 months ago. Claude pulled it up instantly and drafted a full email reply in under 20 minutes using context it had been building for months across 6,000+ sessions. Every session adds to the last one and the system now saves him 2–3 hours every single day - 30–90 minutes from automated morning briefing alone, before he even opens his laptop. repo: The repo is free. The advantage it creates is not.

Noisy

23,679 views • 2 months ago

Anthropic and Andrew Ng built an agent that uses 90% fewer tokens from scratch: they dropped the entire book of Frankenstein into a prompt - 108,000 tokens asked one question the input dropped from 108,000 tokens to 11 here's how: step 1 → put everything that never changes at the top - tools, then system, then docs step 2 → mark where the static part ends. everything above it gets cached step 3 → one stray space breaks it - and you pay full price again step 4 → the cache dies in 5 min - every read resets the clock step 5 → cached tokens don't count against your rate limits. free headroom most people never touch this - it pays for itself on day one watch & bookmark - this 1-hour brilliant course ↓

Anthropic and Andrew Ng built an agent that uses 90% fewer tokens from scratch: they dropped the entire book of Frankenstein into a prompt - 108,000 tokens asked one question the input dropped from 108,000 tokens to 11 here's how: step 1 → put everything that never changes at the top - tools, then system, then docs step 2 → mark where the static part ends. everything above it gets cached step 3 → one stray space breaks it - and you pay full price again step 4 → the cache dies in 5 min - every read resets the clock step 5 → cached tokens don't count against your rate limits. free headroom most people never touch this - it pays for itself on day one watch & bookmark - this 1-hour brilliant course ↓

codila

542,204 views • 25 days ago

$THIS MIGHT BE THE #1 OPEN-SOURCE REPO FOR CLAUDE CODE RIGHT NOW. IT GIVES CLAUDE A MEMORY AND SLASHES YOUR TOKEN COST ON EVERY QUESTION The repo is safishamsi/graphify, a free open-source skill that turns any codebase into a knowledge graph Claude Code can read instantly. Instead of grepping through your files every session, Claude gets a map of how everything connects The problem it fixes: Every time you ask Claude Code about a big repo, it does the same thing, greps through dozens of files like a brute-force Ctrl+F, blows through your context window, and sometimes still misses the answer hiding in a file nobody searched. Claude Code has no memory of how your project is structured. Every session starts from zero What it does: It maps your entire codebase into a knowledge graph, capturing not just which files exist, but which functions depend on which, which modules are central, and which files cluster around the same concern. Claude queries the map instead of scanning files How it works, three passes: 1. Code structure, free and local. Tree-sitter parses your files and pulls out classes, functions, imports and call graphs. No LLM, no tokens, just your actual code mapped deterministically 2. Audio and video, if you have them. Transcribed locally and folded into the graph 3. Docs, papers, images. Here an LLM does semantic analysis, figuring out what each document means and where it fits. Only the meaning gets sent up, never your raw source It saves you money: Normally a question about a big repo makes Claude spawn explore agents that scan file after file, eating your context window and your token budget before you get an answer. With the graph already built, Claude queries the map instead of re-reading the codebase every time. Same answer, a fraction of the tokens. The graph only gets built once, then a hook rebuilds it after each commit for free, so you never pay that scanning cost again. The bigger the repo, the bigger the gap The best parts: it's a skill, so once installed Claude knows when to use it without you memorizing commands. It works on non-code folders too, point it at docs or notes and it can spin up an Obsidian vault How to add it to your Claude: 1. Install Claude Code if you haven't: npm install -g Paul Jankura-ai/claude-code 2. Add the skill: claude skill add safishamsi/graphify 3. Open your project folder and run /graphify . to build the graph 4. Optional, make it automatic: graphify hook install so the graph rebuilds after every commit That's it. Ask Claude about your repo and it reads the map instead of burning tokens on a file hunt Bookmark this$

THIS MIGHT BE THE #1 OPEN-SOURCE REPO FOR CLAUDE CODE RIGHT NOW. IT GIVES CLAUDE A MEMORY AND SLASHES YOUR TOKEN COST ON EVERY QUESTION The repo is safishamsi/graphify, a free open-source skill that turns any codebase into a knowledge graph Claude Code can read instantly. Instead of grepping through your files every session, Claude gets a map of how everything connects The problem it fixes: Every time you ask Claude Code about a big repo, it does the same thing, greps through dozens of files like a brute-force Ctrl+F, blows through your context window, and sometimes still misses the answer hiding in a file nobody searched. Claude Code has no memory of how your project is structured. Every session starts from zero What it does: It maps your entire codebase into a knowledge graph, capturing not just which files exist, but which functions depend on which, which modules are central, and which files cluster around the same concern. Claude queries the map instead of scanning files How it works, three passes: 1. Code structure, free and local. Tree-sitter parses your files and pulls out classes, functions, imports and call graphs. No LLM, no tokens, just your actual code mapped deterministically 2. Audio and video, if you have them. Transcribed locally and folded into the graph 3. Docs, papers, images. Here an LLM does semantic analysis, figuring out what each document means and where it fits. Only the meaning gets sent up, never your raw source It saves you money: Normally a question about a big repo makes Claude spawn explore agents that scan file after file, eating your context window and your token budget before you get an answer. With the graph already built, Claude queries the map instead of re-reading the codebase every time. Same answer, a fraction of the tokens. The graph only gets built once, then a hook rebuilds it after each commit for free, so you never pay that scanning cost again. The bigger the repo, the bigger the gap The best parts: it's a skill, so once installed Claude knows when to use it without you memorizing commands. It works on non-code folders too, point it at docs or notes and it can spin up an Obsidian vault How to add it to your Claude: 1. Install Claude Code if you haven't: npm install -g Paul Jankura-ai/claude-code 2. Add the skill: claude skill add safishamsi/graphify 3. Open your project folder and run /graphify . to build the graph 4. Optional, make it automatic: graphify hook install so the graph rebuilds after every commit That's it. Ask Claude about your repo and it reads the map instead of burning tokens on a file hunt Bookmark this

Yarchi

55,345 views • 1 month ago

most agent memory is one file that grows forever and gets re-read every single turn. it overflows the context, overwrites old facts, loses the thread. that's why your agent feels sharp on day one and lost by week three. Sibyl Memory replaces the pile with a structure: → leaner context, lower token cost. a bloated agent file and flat memory get re-read in full every message. Sibyl Memory keeps a light hot layer and retrieves the rest on demand, so each turn carries only the tokens that matter. → relations are first-class. people link to projects, projects to deals, deals to the decisions behind them. ask about one partner and the agent surfaces the whole connected web around them. that relational context is what lets it help run a company. → one source of truth per fact, person, and project. no duplicates, no silent overwrites. works with hermes, claude code, and codex. beta open.

most agent memory is one file that grows forever and gets re-read every single turn. it overflows the context, overwrites old facts, loses the thread. that's why your agent feels sharp on day one and lost by week three. Sibyl Memory replaces the pile with a structure: → leaner context, lower token cost. a bloated agent file and flat memory get re-read in full every message. Sibyl Memory keeps a light hot layer and retrieves the rest on demand, so each turn carries only the tokens that matter. → relations are first-class. people link to projects, projects to deals, deals to the decisions behind them. ask about one partner and the agent surfaces the whole connected web around them. that relational context is what lets it help run a company. → one source of truth per fact, person, and project. no duplicates, no silent overwrites. works with hermes, claude code, and codex. beta open.

SIBYL

81,361 views • 2 months ago

this video is the CLEAREST explanation of how claude skills + AI agents work and how to use them most people set up an AI agent and wonder why it keeps disappointing them. the context window is everything context is what the model assembles before it takes any action. think of it like everything the agent needs to read before it does anything. the quality of what goes in determines the quality of what comes out. the models are genuinely really good right now. claude and gpt are exceptional. the variable is almost always the context you give them. 1. agent.md files are mostly unnecessary every single line you put in an agent.md file gets added to every single conversation you have with your agent. a 1000 line file is around 7000 tokens burning on every run. the model already knows to use react. it can read your codebase. save the agent.md for proprietary information specific to your company that the model genuinely cannot know on its own. 2. skills are the actual unlock a skill.md file works differently. what loads into context is only the name and description, around 50 tokens. the full instructions only appear when the agent recognizes it needs that skill. so instead of 7000 tokens on every run you have 50. and the agent stays sharp because the context window stays lean. the closer you get to filling the context window the worse the agent performs, same way you perform worse when someone dumps 10 things on you at once. 3. here is how to actually build a skill the right way most people identify a workflow and immediately try to write the skill. what you want to do instead is run the workflow by hand with the agent first. walk it through every single step. tell it what to check, what good looks like, what bad looks like. correct it in real time. once you have had a full successful run from start to finish, tell the agent to review everything it just did and write the skill itself. it writes a better skill than you will because it has the full context of what actually worked in practice not in theory. 4. recursively building skills is how you go from frustrated to reliable when the skill breaks, and it will break, ask the agent exactly why it failed. it will tell you specifically what went wrong. fix it together in that same conversation. then tell it to update the skill file so that failure mode never happens again. ross mike did this five times with his youtube report generator. it now pulls from eight different data sources and runs flawlessly every single time without him touching it. 5. sub agents are something you earn not something you set up on day one start with one agent. build one workflow. turn it into one skill. once that works add another. ross mike has five sub agents now covering marketing, business, personal and more. it took months to get there and every single one exists because a workflow proved it deserved to exist. the people who set up 15 sub agents on day one and wonder why nothing works skipped all the steps that make the thing actually run. 6. your workflow is the thing the model cannot get anywhere else the model has been trained on everything. it knows more than you about most things. what it does not have is your specific process, your taste, your way of doing things. that is what skills capture. that is what makes your agent actually useful versus a generic one. downloading someone else's skill means downloading their context onto your setup and it will not work the way you want it to because it was never built around how you work. this is the clearest explanation of how agents actually work i have heard. Micky runs this stuff every single day and the results show it. full episode is now live on The Startup Ideas Podcast (SIP) 🧃 where you get your pods people charge for this sorta stuff i give away the sauce for free i just want you to win watch

this video is the CLEAREST explanation of how claude skills + AI agents work and how to use them most people set up an AI agent and wonder why it keeps disappointing them. the context window is everything context is what the model assembles before it takes any action. think of it like everything the agent needs to read before it does anything. the quality of what goes in determines the quality of what comes out. the models are genuinely really good right now. claude and gpt are exceptional. the variable is almost always the context you give them. 1. agent.md files are mostly unnecessary every single line you put in an agent.md file gets added to every single conversation you have with your agent. a 1000 line file is around 7000 tokens burning on every run. the model already knows to use react. it can read your codebase. save the agent.md for proprietary information specific to your company that the model genuinely cannot know on its own. 2. skills are the actual unlock a skill.md file works differently. what loads into context is only the name and description, around 50 tokens. the full instructions only appear when the agent recognizes it needs that skill. so instead of 7000 tokens on every run you have 50. and the agent stays sharp because the context window stays lean. the closer you get to filling the context window the worse the agent performs, same way you perform worse when someone dumps 10 things on you at once. 3. here is how to actually build a skill the right way most people identify a workflow and immediately try to write the skill. what you want to do instead is run the workflow by hand with the agent first. walk it through every single step. tell it what to check, what good looks like, what bad looks like. correct it in real time. once you have had a full successful run from start to finish, tell the agent to review everything it just did and write the skill itself. it writes a better skill than you will because it has the full context of what actually worked in practice not in theory. 4. recursively building skills is how you go from frustrated to reliable when the skill breaks, and it will break, ask the agent exactly why it failed. it will tell you specifically what went wrong. fix it together in that same conversation. then tell it to update the skill file so that failure mode never happens again. ross mike did this five times with his youtube report generator. it now pulls from eight different data sources and runs flawlessly every single time without him touching it. 5. sub agents are something you earn not something you set up on day one start with one agent. build one workflow. turn it into one skill. once that works add another. ross mike has five sub agents now covering marketing, business, personal and more. it took months to get there and every single one exists because a workflow proved it deserved to exist. the people who set up 15 sub agents on day one and wonder why nothing works skipped all the steps that make the thing actually run. 6. your workflow is the thing the model cannot get anywhere else the model has been trained on everything. it knows more than you about most things. what it does not have is your specific process, your taste, your way of doing things. that is what skills capture. that is what makes your agent actually useful versus a generic one. downloading someone else's skill means downloading their context onto your setup and it will not work the way you want it to because it was never built around how you work. this is the clearest explanation of how agents actually work i have heard. Micky runs this stuff every single day and the results show it. full episode is now live on The Startup Ideas Podcast (SIP) 🧃 where you get your pods people charge for this sorta stuff i give away the sauce for free i just want you to win watch

GREG ISENBERG

193,219 views • 4 months ago

Peter Steinberger Quietly Started A Shift That Makes Claude 35x Cheaper To Run The way Claude Code talks to tools matters more than people think. Most use MCP, and it quietly eats tokens, it loads everything into context every time, even tools you never call. One benchmark: MCP burned 35x more tokens than a CLI on the same task. Peter Steinberger, the OpenClaw guy, got annoyed by this and started building lean CLIs himself. That kicked off a tool called Printing Press. You point Claude Code at any website, even ones with no API like ESPN or Craigslist, and it builds a small command tool for it in about 10 minutes. In one demo, a request pulled 132,000 tokens of raw data, but the tool processed it locally and only handed Claude a 2,000-token summary. The rest never touched the context window. It also comes with ~50 ready-made tools you can grab right away. To start, point Claude Code at the links from and ask it to set it up. Bookmark this.

Peter Steinberger Quietly Started A Shift That Makes Claude 35x Cheaper To Run The way Claude Code talks to tools matters more than people think. Most use MCP, and it quietly eats tokens, it loads everything into context every time, even tools you never call. One benchmark: MCP burned 35x more tokens than a CLI on the same task. Peter Steinberger, the OpenClaw guy, got annoyed by this and started building lean CLIs himself. That kicked off a tool called Printing Press. You point Claude Code at any website, even ones with no API like ESPN or Craigslist, and it builds a small command tool for it in about 10 minutes. In one demo, a request pulled 132,000 tokens of raw data, but the tool processed it locally and only handed Claude a 2,000-token summary. The rest never touched the context window. It also comes with ~50 ready-made tools you can grab right away. To start, point Claude Code at the links from and ask it to set it up. Bookmark this.

Ridark

32,351 views • 1 month ago

Anthropic and Andrew Ng built an agent that uses 90% fewer tokens from scratch: they dropped the entire book of Frankenstein into a prompt - 108,000 tokens asked one question the input dropped from 108,000 tokens to 11 here's how: step 1 → order matters: tools first, then system prompt, then your docs. anything that changes goes last step 2 → drop the breakpoint after the last stable block. everything above it gets stored step 3 → the match has to be byte-perfect. one extra space and you're back to full price step 4 → 5 minutes of idle and it's gone. but every hit restarts the timer step 5 → cache reads don't touch your rate limit. that's throughput you're not paying for 99% people never configure this - one afternoon and it pays for itself save this - the full 1-hour course is right below ↓

Anthropic and Andrew Ng built an agent that uses 90% fewer tokens from scratch: they dropped the entire book of Frankenstein into a prompt - 108,000 tokens asked one question the input dropped from 108,000 tokens to 11 here's how: step 1 → order matters: tools first, then system prompt, then your docs. anything that changes goes last step 2 → drop the breakpoint after the last stable block. everything above it gets stored step 3 → the match has to be byte-perfect. one extra space and you're back to full price step 4 → 5 minutes of idle and it's gone. but every hit restarts the timer step 5 → cache reads don't touch your rate limit. that's throughput you're not paying for 99% people never configure this - one afternoon and it pays for itself save this - the full 1-hour course is right below ↓

codila

81,033 views • 2 days ago

THIS IS HOW A SENIOR ENGINEER ACTUALLY SCALES THEMSELVES WITH CLAUDE CODE the biggest change with AI isn't coding faster. it's where you actually spend your time now. more detailed prompts, more code review, more planning, less typing, etc. here's the workflow: this guy has been shipping code since the days of cgi and perl. he uses a compound engineering plugin that runs 5 separate agents on every task. one brainstorms, one plans the technical implementation, one executes, one reviews, one checks different verticals. every step is documented in markdown files. it's slow and way more waiting. but the output quality is way higher because each agent is focused on one thing. then the REAL multiplier is in git worktrees if Claude Code made you 10x faster, worktrees multiplies that again depending on how many agents you can manage in parallel his team runs 4-8 Claude Code sessions at the same time across different worktrees with each one working on a separate task. the skill is managing multiple AI agents in parallel without losing track, that's the next evolution of engineering

THIS IS HOW A SENIOR ENGINEER ACTUALLY SCALES THEMSELVES WITH CLAUDE CODE the biggest change with AI isn't coding faster. it's where you actually spend your time now. more detailed prompts, more code review, more planning, less typing, etc. here's the workflow: this guy has been shipping code since the days of cgi and perl. he uses a compound engineering plugin that runs 5 separate agents on every task. one brainstorms, one plans the technical implementation, one executes, one reviews, one checks different verticals. every step is documented in markdown files. it's slow and way more waiting. but the output quality is way higher because each agent is focused on one thing. then the REAL multiplier is in git worktrees if Claude Code made you 10x faster, worktrees multiplies that again depending on how many agents you can manage in parallel his team runs 4-8 Claude Code sessions at the same time across different worktrees with each one working on a separate task. the skill is managing multiple AI agents in parallel without losing track, that's the next evolution of engineering

Om Patel

159,441 views • 4 months ago

Claude Code Head invented a word for what most config files aren't doing: tokenmaxxing. 90% of it lives in one place: ~/.claude/settings.json. 125 keys in v2.1.105. official docs cover 40. 4 of mine aren't anywhere in the docs. i audited the whole file - 18 settings actually decide what your month costs. > enabledPlugins - mine had 14 on, 4 actually used. each idle plugin loads hooks and SKILL.md at session start. > mcpServers.enabled - false instead of delete. 9 unused servers cost ~30K tokens of schema every session. > cache_control breakpoint at the static/dynamic boundary moved my Opus month from $340 to $87. tokenmaxxing isn't a vibe. it's a file with 125 keys most people never open. watch this tonight instead of the new Rick and Morty drop.

Claude Code Head invented a word for what most config files aren't doing: tokenmaxxing. 90% of it lives in one place: ~/.claude/settings.json. 125 keys in v2.1.105. official docs cover 40. 4 of mine aren't anywhere in the docs. i audited the whole file - 18 settings actually decide what your month costs. > enabledPlugins - mine had 14 on, 4 actually used. each idle plugin loads hooks and SKILL.md at session start. > mcpServers.enabled - false instead of delete. 9 unused servers cost ~30K tokens of schema every session. > cache_control breakpoint at the static/dynamic boundary moved my Opus month from $340 to $87. tokenmaxxing isn't a vibe. it's a file with 125 keys most people never open. watch this tonight instead of the new Rick and Morty drop.

Mnimiy

83,403 views • 2 months ago

$Tom Crawshaw has been building automations for 9 years with $25 million in client revenue to his name. He just walked me through his Claude Code content system that's generating millions of views and tens of thousands of followers. Here's what I learned: 1. Skills beat Projects in Claude. Projects load every context file on every message and burn your token window. Skills work like a book where the LLM reads the table of contents and pulls only the chapter it needs for the job. Same context, fraction of the tokens. 2. He has a /content-create slash command that runs the entire pipeline. Voice profile, copywriting principles, hook generation, image direction. One command. He doesn't write anything from scratch anymore. 3. His voice profile auto-updates weekly. He wrote a script that hits the X API every 7 days, pulls his top-performing posts by engagement, and rewrites his voice profile based on what's actually working. The profile evolves on its own. 4. He distilled a master copywriter's entire body of work into a single markdown file. Grabbed every Alen Sultanic post he could find, dropped it into Claude, asked for the core principles, fed it into the skill. Now every post he writes runs through those principles automatically. 5. The hook generator scores 16 hooks per post against 7 criteria. Curiosity loops, specificity, sensory, credibility, voice match, and a couple of others. He never picks the #1 by default. Sometimes he splices the first line of one hook with the body of another. The taste is still his. 6. /insights is a native Claude Code command nobody talks about. It analyzes every session you've ever run and produces a full report on your usage patterns, where things break down, and prompts you can paste back into Claude to fix them. I had never heard of it. I'm running it tonight. 7. He spends most of his time on hooks and images. If those two suck, the body copy doesn't matter. Nobody reads it. 8. Image generation is never one-shot. He keeps a folder of reference images that have worked before and feeds them into Nano Banana/GPT Image 2 every time. Then he takes the 80-90% output and finishes it in Canva using "magic grab" to move logos and clean up text. Last mile is human every time. 9. The humanizer step is non-negotiable. Strip em dashes. Kill the "it's not X, it's Y" pattern. Cut the triple negatives. Cut "no fluff." He still has to remind Claude mid-session because it drifts. If you're not auditing for AI tells, you're shipping slop. 10. Wisprflow is the most important tool in his stack. Not a content tool. An everything tool. His test for whether you should be using it: do you talk faster than you type? You do. Everyone does. Bonus fact he dropped: QWERTY was designed 200 years ago to slow typing down so old typewriters wouldn't jam. We've been carrying that forward ever since. Voice is finally undoing it. This was an inside look at how a serious operator Tom is using Claude Code to run a content engine. The good, the bad, the iteration, all of it. I hope you enjoy this one as much as I did. Go watch it.$

Tom Crawshaw has been building automations for 9 years with $25 million in client revenue to his name. He just walked me through his Claude Code content system that's generating millions of views and tens of thousands of followers. Here's what I learned: 1. Skills beat Projects in Claude. Projects load every context file on every message and burn your token window. Skills work like a book where the LLM reads the table of contents and pulls only the chapter it needs for the job. Same context, fraction of the tokens. 2. He has a /content-create slash command that runs the entire pipeline. Voice profile, copywriting principles, hook generation, image direction. One command. He doesn't write anything from scratch anymore. 3. His voice profile auto-updates weekly. He wrote a script that hits the X API every 7 days, pulls his top-performing posts by engagement, and rewrites his voice profile based on what's actually working. The profile evolves on its own. 4. He distilled a master copywriter's entire body of work into a single markdown file. Grabbed every Alen Sultanic post he could find, dropped it into Claude, asked for the core principles, fed it into the skill. Now every post he writes runs through those principles automatically. 5. The hook generator scores 16 hooks per post against 7 criteria. Curiosity loops, specificity, sensory, credibility, voice match, and a couple of others. He never picks the #1 by default. Sometimes he splices the first line of one hook with the body of another. The taste is still his. 6. /insights is a native Claude Code command nobody talks about. It analyzes every session you've ever run and produces a full report on your usage patterns, where things break down, and prompts you can paste back into Claude to fix them. I had never heard of it. I'm running it tonight. 7. He spends most of his time on hooks and images. If those two suck, the body copy doesn't matter. Nobody reads it. 8. Image generation is never one-shot. He keeps a folder of reference images that have worked before and feeds them into Nano Banana/GPT Image 2 every time. Then he takes the 80-90% output and finishes it in Canva using "magic grab" to move logos and clean up text. Last mile is human every time. 9. The humanizer step is non-negotiable. Strip em dashes. Kill the "it's not X, it's Y" pattern. Cut the triple negatives. Cut "no fluff." He still has to remind Claude mid-session because it drifts. If you're not auditing for AI tells, you're shipping slop. 10. Wisprflow is the most important tool in his stack. Not a content tool. An everything tool. His test for whether you should be using it: do you talk faster than you type? You do. Everyone does. Bonus fact he dropped: QWERTY was designed 200 years ago to slow typing down so old typewriters wouldn't jam. We've been carrying that forward ever since. Voice is finally undoing it. This was an inside look at how a serious operator Tom is using Claude Code to run a content engine. The good, the bad, the iteration, all of it. I hope you enjoy this one as much as I did. Go watch it.

Corey Ganim

36,120 views • 3 months ago

THIS GUY TURNED ONE OBSIDIAN VAULT INTO A SHARED BRAIN FOR 5 CLAUDE AGENTS BUILT TO HANDLE 80% OF A SOLO DEVELOPER’S WORK. 00:03 he zooms out and hundreds of connected notes appear on screen. it is the perfect visual for Claude working from one shared project memory instead of starting from zero every session. most developers give Claude 1 file and 1 prompt. this setup gives it the architecture, coding rules, tests, client briefs, and previous decisions before it writes anything. obsidian becomes the memory layer, storing everything in markdown files Claude can read alongside the codebase. every agent works from the same context instead of repeating the same questions. then 5 specialists split the job. one researches, one plans, one codes, one reviews, and one deploys while the developer only controls the important decisions. one vault, 5 agents, 2 model tiers, and a $50 to $150 monthly stack can turn Claude from a chatbot into a small development team.

THIS GUY TURNED ONE OBSIDIAN VAULT INTO A SHARED BRAIN FOR 5 CLAUDE AGENTS BUILT TO HANDLE 80% OF A SOLO DEVELOPER’S WORK. 00:03 he zooms out and hundreds of connected notes appear on screen. it is the perfect visual for Claude working from one shared project memory instead of starting from zero every session. most developers give Claude 1 file and 1 prompt. this setup gives it the architecture, coding rules, tests, client briefs, and previous decisions before it writes anything. obsidian becomes the memory layer, storing everything in markdown files Claude can read alongside the codebase. every agent works from the same context instead of repeating the same questions. then 5 specialists split the job. one researches, one plans, one codes, one reviews, and one deploys while the developer only controls the important decisions. one vault, 5 agents, 2 model tiers, and a $50 to $150 monthly stack can turn Claude from a chatbot into a small development team.

Gipp 🦅

10,776 views • 1 month ago

how to use Google's NEW open source Design.md + AI Skills to make your startup look like a $100 million company in 1 hour: 1. Design.md is an open source file from Google that captures the soul of a design. Typography, colors, spacing, all in one markdown file. You attach it to your prompt and your agent builds beautiful things every time. 2. Think of it this way. The HTML is the finished dish. The design.md is the recipe. The skills are the ingredients. Put them together and everything you build looks consistent and professional. 3. Don't create a design system from scratch. Find a brand you love. Linear, Stripe, Vercel, whatever resonates. Study it. Use ChatGPT or Claude to help you extract the design language into your own design.md file. 4. Build skills on top of your design.md. A landing page skill. A mobile app skill. A motion design skill. A slide deck skill. Each one references the same design.md so everything looks like it came from the same designer. 5. The biggest mistake people make: they nail one screen and then everything else looks generic. Design.md solves this. One file keeps every page, every format, every medium consistent. 6. Use it across everything. Your landing page. Your app. Your pitch deck. Your promo videos. Same DNA. Same taste. Same system. That's what separates a startup that looks real from one that looks vibe-coded. 7. Build a second brain for design inspiration. When you see something beautiful in the real world or online, capture it. Save it. When you're building something new, reference it. Taste is developed, not downloaded. 8. It's obvious but the difference between a product people trust and a product people bounce from is how it looks and feels. Design.md gives you that edge. you can watch below shoutout to Meng To for coming on The Startup Ideas Podcast (SIP) 🧃 and walking through his full workflow. if you want to use AI to actually build gorgeous designs, you'll want to use see this. watch

GREG ISENBERG

507,989 views • 3 months ago

andrej karpathy spent two hours teaching one thing: tokens are the atom of llms. tokenization is at the heart of every llm weirdness you've ever debugged. [watch the 15-min clip below. then run the 7-day playbook] ↓ save this before everyone copies it learn how the tokenizer works. understand how your llm actually consumes input. then run the engineering roadmap that took one production agent from $4,800/mo to $620/mo in 7 days. 87% reduction. no model swap. no framework migration. no quality drop on the eval set. token cost in 2026 is an engineering discipline. every line of your system prompt is rent you pay forever. what was eating the budget: → a single forgotten cron job ate 47% of one team's bill. they turned it off on a tuesday and the bill dropped before they wrote any optimization code. → anthropic ships a 90% discount on cache reads. one config line, cache_control ephemeral, break-even after one hit. most teams cache the volatile parts of the prompt and watch their hit rate sit at 12%. → one production agent went from 14,500 tokens of context overhead per turn to 850. a 94% drop. output quality held within 2% of the uncompressed baseline. → 60% of agent calls are haiku-tier work running on opus rates. classify the task first. pick the model second. → retry loops are the silent killer. no MAX_STEPS bound, one bad search query, $14 burned in a single session. one team traced 38% of their bill to this single pattern. karpathy gave you the atom. the playbook below gives you the harness. watch the lecture. read the playbook ↓

andrej karpathy spent two hours teaching one thing: tokens are the atom of llms. tokenization is at the heart of every llm weirdness you've ever debugged. [watch the 15-min clip below. then run the 7-day playbook] ↓ save this before everyone copies it learn how the tokenizer works. understand how your llm actually consumes input. then run the engineering roadmap that took one production agent from $4,800/mo to $620/mo in 7 days. 87% reduction. no model swap. no framework migration. no quality drop on the eval set. token cost in 2026 is an engineering discipline. every line of your system prompt is rent you pay forever. what was eating the budget: → a single forgotten cron job ate 47% of one team's bill. they turned it off on a tuesday and the bill dropped before they wrote any optimization code. → anthropic ships a 90% discount on cache reads. one config line, cache_control ephemeral, break-even after one hit. most teams cache the volatile parts of the prompt and watch their hit rate sit at 12%. → one production agent went from 14,500 tokens of context overhead per turn to 850. a 94% drop. output quality held within 2% of the uncompressed baseline. → 60% of agent calls are haiku-tier work running on opus rates. classify the task first. pick the model second. → retry loops are the silent killer. no MAX_STEPS bound, one bad search query, $14 burned in a single session. one team traced 38% of their bill to this single pattern. karpathy gave you the atom. the playbook below gives you the harness. watch the lecture. read the playbook ↓

Rohit

73,258 views • 2 months ago

I plugged MiniMax (official) M3 into Claude Code in under 5 minutes and gave it a real task. most models need hand-holding every few steps. M3 ran 147 iterations on its own. what it actually did: → took a broken CUDA kernel → ran it for 24 hours, autonomously → pushed hardware efficiency from 7.6% to 71.3% → shipped a 9.4x speedup I didn't touch it once across that whole window. the rest of the spec sheet: → 1M token context window → native multimodal from day one → fully open weights pricing: → $20/month for 1.7B tokens → $50/month for 5.1B tokens → text, image, speech, and music all share one pool usually with this kind of capability the bill catches up fast. M3 keeps both lines low at the same time. that's the part that surprised me. and if this is what M3 does at launch, the full rollout is going to be one to watch. what would you build with 1M context? #MiniMax #MiniMaxM3 #ClaudeCode #LLM

I plugged MiniMax (official) M3 into Claude Code in under 5 minutes and gave it a real task. most models need hand-holding every few steps. M3 ran 147 iterations on its own. what it actually did: → took a broken CUDA kernel → ran it for 24 hours, autonomously → pushed hardware efficiency from 7.6% to 71.3% → shipped a 9.4x speedup I didn't touch it once across that whole window. the rest of the spec sheet: → 1M token context window → native multimodal from day one → fully open weights pricing: → $20/month for 1.7B tokens → $50/month for 5.1B tokens → text, image, speech, and music all share one pool usually with this kind of capability the bill catches up fast. M3 keeps both lines low at the same time. that's the part that surprised me. and if this is what M3 does at launch, the full rollout is going to be one to watch. what would you build with 1M context? #MiniMax #MiniMaxM3 #ClaudeCode #LLM

Nelly;

17,576 views • 2 months ago

I cant believe this guy just made a permanent solution to context bloat and open sourced it all! when we tested this tool (Context+) for solving an issue on the OpenCode repository, the agent using this tool used ~6.5k fewer tokens, found the code and fixed it in half the time! the results were surprising: 6 to 10k tokens saved per prompt, completed task in ~2 minutes while the agent running without the tool took ~4 mins for the same and got stuck in loops bro built an entire beast by using all the modern tools that we could think of: undo trees, semantic search by meaning (by haskellforall), advanced refactoring, blast radius, advanced file context trees, restore points... i can keep going on semantic code search and context trees are the future of agentic coding and this tool proves it the feature i loved the most is semantic search and how it gets things done 2x faster with least possible tokens it makes an agent that actually knows what it’s doing and not just guessing, it makes meaning from your code similar to RAG. if you aren't optimizing your context, you are just burning money the developer says this tool is still under development, it can have unexpected behavior and the docs need updates but the video shows the reality of how fast it can be github: get here:

I cant believe this guy just made a permanent solution to context bloat and open sourced it all! when we tested this tool (Context+) for solving an issue on the OpenCode repository, the agent using this tool used ~6.5k fewer tokens, found the code and fixed it in half the time! the results were surprising: 6 to 10k tokens saved per prompt, completed task in ~2 minutes while the agent running without the tool took ~4 mins for the same and got stuck in loops bro built an entire beast by using all the modern tools that we could think of: undo trees, semantic search by meaning (by haskellforall), advanced refactoring, blast radius, advanced file context trees, restore points... i can keep going on semantic code search and context trees are the future of agentic coding and this tool proves it the feature i loved the most is semantic search and how it gets things done 2x faster with least possible tokens it makes an agent that actually knows what it’s doing and not just guessing, it makes meaning from your code similar to RAG. if you aren't optimizing your context, you are just burning money the developer says this tool is still under development, it can have unexpected behavior and the docs need updates but the video shows the reality of how fast it can be github: get here:

forloop

226,054 views • 5 months ago

A 26-YEAR-OLD AI ENGINEER WHOSE BIRMINGHAM THESIS OUTPERFORMED GOOGLE SCHOLAR BY 50% JUST SHIPPED GRAPHIFY: ONE COMMAND TURNS ANY FOLDER INTO A CLAUDE CODE SECOND BRAIN Safi Shamsi built Graphify 48 hours after Karpathy posted his LLM wiki idea. It turns any folder, codebase, docs, PDFs, into a knowledge graph Claude reads instead of grepping. Up to 43x fewer tokens per query. The trick almost nobody is using yet: one flag exports the entire graph as a fully-linked Obsidian vault. Repo: /safishamsi/graphify Setup, end to end: 1. Install: uv tool install graphifyy (or pipx install graphifyy). Verify with graphify --version. 2. Install the skill in Claude Code: graphify claude install. This wires Graphify into Claude Code so you can call it as a skill. 3. Open the folder you want mapped in Claude Code. In the terminal: graphify . It extracts every concept and builds the graph in graphify-out/. 4. Export to Obsidian: graphify . --obsidian. Writes one note per concept, every relationship as a wikilink, every node linked back to its source. 5. Open the new vault in Obsidian (Manage vaults → Open folder as a vault), or drag it as a subfolder into your existing one. That's it. Your Claude Code instance now has a navigable map of the codebase that loads instantly instead of re-reading files every session. Full step-by-step build of the Claude + Obsidian second brain in the article below. Bookmark this

A 26-YEAR-OLD AI ENGINEER WHOSE BIRMINGHAM THESIS OUTPERFORMED GOOGLE SCHOLAR BY 50% JUST SHIPPED GRAPHIFY: ONE COMMAND TURNS ANY FOLDER INTO A CLAUDE CODE SECOND BRAIN Safi Shamsi built Graphify 48 hours after Karpathy posted his LLM wiki idea. It turns any folder, codebase, docs, PDFs, into a knowledge graph Claude reads instead of grepping. Up to 43x fewer tokens per query. The trick almost nobody is using yet: one flag exports the entire graph as a fully-linked Obsidian vault. Repo: /safishamsi/graphify Setup, end to end: 1. Install: uv tool install graphifyy (or pipx install graphifyy). Verify with graphify --version. 2. Install the skill in Claude Code: graphify claude install. This wires Graphify into Claude Code so you can call it as a skill. 3. Open the folder you want mapped in Claude Code. In the terminal: graphify . It extracts every concept and builds the graph in graphify-out/. 4. Export to Obsidian: graphify . --obsidian. Writes one note per concept, every relationship as a wikilink, every node linked back to its source. 5. Open the new vault in Obsidian (Manage vaults → Open folder as a vault), or drag it as a subfolder into your existing one. That's it. Your Claude Code instance now has a navigable map of the codebase that loads instantly instead of re-reading files every session. Full step-by-step build of the Claude + Obsidian second brain in the article below. Bookmark this

Yarchi

140,729 views • 1 month ago

Claude Code creator: "100% of our pull requests at Anrtopic are run by Claude Code. 80–90% of code review too. The feature I’m using the most today is /loops. I’m not prompting Claude anymore - I’m building loops" in this interview he breaks down exactly how a system changes everything: - the memory and context features that turn Claude into a second brain - the knowledge architecture most users don't know Claude can build - the integration layer that connects Claude to your actual workflow - why typing one question at a time is the most expensive way to use Claude if you've been using Claude for months and still start every conversation from scratch with zero context, you don't have a Claude problem. you have a system problem instead of another show tonight, watch this make sure to bookmark it before it gets lost in your feed full guide in the article below

Claude Code creator: "100% of our pull requests at Anrtopic are run by Claude Code. 80–90% of code review too. The feature I’m using the most today is /loops. I’m not prompting Claude anymore - I’m building loops" in this interview he breaks down exactly how a system changes everything: - the memory and context features that turn Claude into a second brain - the knowledge architecture most users don't know Claude can build - the integration layer that connects Claude to your actual workflow - why typing one question at a time is the most expensive way to use Claude if you've been using Claude for months and still start every conversation from scratch with zero context, you don't have a Claude problem. you have a system problem instead of another show tonight, watch this make sure to bookmark it before it gets lost in your feed full guide in the article below

Khairallah AL-Awady

39,243 views • 1 month ago

Anthropic shipped 125 settings for Claude The official docs cover 40 One developer found the other 85 and his API bill dropped from $340 to $87 - not by using a cheaper model - not by writing shorter prompts just by moving one line in a config file to the right place > memory scoped per project → past clients never bleed into new work > Extended Thinking on Light by default → 18–25% fewer Opus tokens in week one > cache_control moved to the right line → the fix that turned a $340 bill into $87 > plugins and MCP servers toggled off when idle → saved 25–40K tokens per session > per-project model override → Haiku for docs, Sonnet for infra, Opus only where it matters most Claude users are running a $100/month tool at 30% of its actual capability

Anthropic shipped 125 settings for Claude The official docs cover 40 One developer found the other 85 and his API bill dropped from $340 to $87 - not by using a cheaper model - not by writing shorter prompts just by moving one line in a config file to the right place > memory scoped per project → past clients never bleed into new work > Extended Thinking on Light by default → 18–25% fewer Opus tokens in week one > cache_control moved to the right line → the fix that turned a $340 bill into $87 > plugins and MCP servers toggled off when idle → saved 25–40K tokens per session > per-project model override → Haiku for docs, Sonnet for infra, Opus only where it matters most Claude users are running a $100/month tool at 30% of its actual capability

Shadow Nick

379,914 views • 2 months ago

context engineering vs graph engineering. every few months the list gets a new word and everyone treats it as a replacement for the last one. these two are not on the same list. one decides what the model sees this turn, the other decides what exists at all. the cleanest way to tell them apart is to ask what a single unit of work looks like. > context engineering is the window the window opens empty, every single time. you assemble what goes in it. the prompt, the docs, the history, the tool results. the assembling is the work. the window only grows. it never shrinks on its own, so eventually something gets dropped. usually from the middle. usually without telling you. then the turn ends and the window is thrown away. not archived, thrown away. the next turn opens empty again and you re-explain what you already explained. good context engineering is knowing what to leave out, not what to pack in. the unit of work is one window. > graph engineering is the structure the same material arrives from the same sources. instead of packing it into a window, you pull entities out of it, resolve the duplicates into one node, and write typed edges between them. nothing here is stored as text you hope to find again. it is stored as a thing with a name and its connections to other things. when the turn ends, the graph is still there. the next turn does not start from zero. it starts by querying what already exists, and the query walks edges instead of guessing at similarity. good graph engineering is deciding what counts as the same thing twice. the unit of work is one relationship. > they are not alternatives the graph is what refills the window. context engineering decides what fits. graph engineering decides what there is to choose from. remove the graph and every session starts blind. remove the context work and the best structure in the world arrives as an unreadable dump. that also tells you which one broke. the answer drifted from what you actually said, or forgot something from this same session. that is the window. the answer is coherent but invents a connection that does not exist, or cannot join two facts it has clearly seen. that is the structure. people debug the prompt because the prompt is the easiest thing to edit. it keeps taking the blame for failures that live a layer down. save this - then read the full breakdown below

context engineering vs graph engineering. every few months the list gets a new word and everyone treats it as a replacement for the last one. these two are not on the same list. one decides what the model sees this turn, the other decides what exists at all. the cleanest way to tell them apart is to ask what a single unit of work looks like. > context engineering is the window the window opens empty, every single time. you assemble what goes in it. the prompt, the docs, the history, the tool results. the assembling is the work. the window only grows. it never shrinks on its own, so eventually something gets dropped. usually from the middle. usually without telling you. then the turn ends and the window is thrown away. not archived, thrown away. the next turn opens empty again and you re-explain what you already explained. good context engineering is knowing what to leave out, not what to pack in. the unit of work is one window. > graph engineering is the structure the same material arrives from the same sources. instead of packing it into a window, you pull entities out of it, resolve the duplicates into one node, and write typed edges between them. nothing here is stored as text you hope to find again. it is stored as a thing with a name and its connections to other things. when the turn ends, the graph is still there. the next turn does not start from zero. it starts by querying what already exists, and the query walks edges instead of guessing at similarity. good graph engineering is deciding what counts as the same thing twice. the unit of work is one relationship. > they are not alternatives the graph is what refills the window. context engineering decides what fits. graph engineering decides what there is to choose from. remove the graph and every session starts blind. remove the context work and the best structure in the world arrives as an unreadable dump. that also tells you which one broke. the answer drifted from what you actually said, or forgot something from this same session. that is the window. the answer is coherent but invents a connection that does not exist, or cannot join two facts it has clearly seen. that is the structure. people debug the prompt because the prompt is the easiest thing to edit. it keeps taking the blame for failures that live a layer down. save this - then read the full breakdown below

Hanako

19,089 views • 9 days ago