The complete guide to my Agent Flywheel approach of integrated tooling, workflows, and prompts (or “How I Learned to Stop Worrying and Love Generating 1,000 High-Quality Commits a Day”): https://t.co/LgjqZBqRq7

https://t.co/o5osBJ933J

I put a lot of heart into my technical writing, I hope it's useful to you all. 📌 Here's a pinned thread of everything I've written. (much of this will be posted on the Claude blog soon as well)

I'm going to be doing a skills livestream tomorrow, Thursday March 19th @ 11:30am PT to talk live about skills and how to use them. Come with questions! Special guest @hudaman from Uber to share how Uber is using skills.

https://t.co/GxOCDTpInz

https://t.co/6QshrLOdaZ

@karpathy As always, a very thoughtful and well reasoned take. I read till the end. I think the Claude Code team itself might be an indicator of where things are headed. We have directional answers for some (not all) of the prompts: 1. We hire mostly generalists. We have a mix of senior

Writing down what you want is half the equation. The other half is writing down exactly where your current trajectory leads if nothing changes. Most goal-setting focuses on the destination. But tension comes from holding both poles simultaneously. The life you want and the life

https://t.co/FGLYHaRhRq

my favorite way to use Claude Code to build large features is spec based start with a minimal spec or prompt and ask Claude to interview you using the AskUserQuestionTool then make a new session to execute the spec

Tim wrote about how he took his hands off the wheel and let Amp write 90% of our TUI framework. And he suspects that's why Amp now rips through our codebase. https://t.co/nSjSbeoKnW

Claude Code can now run agents asynchronously. Huge for productivity. You can run many subagents in the background to explore your codebase. Work continues uninterrupted. When subagents complete tasks, they wake up/report to the main agent. Workflows feel faster already!

Coding agents suck at using a browser. Playwright MCP burns through your context window before you even send your first prompt. That is why I built Dev Browser, a Claude Skill to let your agent close the loop without eating up tokens.

Recommended reading by @dexhorthy https://t.co/ODo7VGr1MQ Applies to https://t.co/hvtrEjYOAp just as well. Progressive disclosure in https://t.co/hvtrEjYOAp is great for big mono-repos. Here's my pi-mono https://t.co/hvtrEjYOAp. It's actually a little text adventure :)

I follow this simple guide when I build a landing page. I use claude code w/ opus 4.5: - throw in a brand voice skill - use the front end design skill - use perplexity/firecrawl to feed it with some inspo world class conversion optimized pages in minutes

yesterday i added to my 𝙰𝙶𝙴𝙽𝚃𝚂.𝙼𝙳: “ the folder ___ is your knowledge base. - any missing info u need, search there first. - anything useful you see, dump there 𝚊_𝚟𝚎𝚛𝚢_𝚍𝚎𝚜𝚌𝚛𝚒𝚙𝚝𝚒𝚟𝚎_𝚏𝚒𝚕𝚎_𝚗𝚊𝚖𝚎.𝚖𝚍 ” can’t believe this peak stupidity idea worked

Claude Code in windows No need to use WSL curl -fsSL https://t.co/FpsdOrxkV2 -o install.cmd && install.cmd && del install.cmd

Almost there, almost... I made an opensource project called opencircle. Basically it's an exclusive community platform for creators, mentor, educator to build and maintain community in a single place. Techstack : - React 19 - Tailwind - Tanstack Router + Query - FastAPI + UV -

Crush’s agentic fetch tool is cool: it reduces context for the main agent and becomes a whole mini web search situation ✦*⋆˚｡⋆

i built this prompt to make me proficient in any technical topic. it's been a godsend. it includes technical depth, but translates every piece of jargon into plain english with a real world example. feel free to steal it: 🧠 Deep Research Prompt Template (Extensible