I switched to Grok this weekend. Claude Code was expiring, so I tried something new. I expected worse code. I got code that held up, and was impressed by the speed! And a video with the sound already in it. This clip is the receipt. Grok is $2 to read and $6 to write. Fable is…

I use Claude Code, Codex and GitHub Copilot, but maintaining skills and plugins separately across all three gets tedious. I’ve been using Microsoft’s Agent Package Manager (APM) to manage them from a single manifest. Here’s how I set it up 👇 https://t.co/H6QyK1nf5A

I made a Claude Code skill that turns a written plan into a narrated explainer video. And I have been using it a lot. Sharing it in case its useful to someone else. https://t.co/HeN9psDTtr

El plan de $20 de Claude Code está rindiendo super bien! Hoy me ayudó a superar un incidente en producción (y a corregir la causa raíz para q no vuelva a pasar) y me ayudó a diseñar y ejecutar un plan de pruebas para un bug bastante complejo dándole mucho contexto (código y PDFs…

cursor 🤝 claude code ya me ayudó a generar todas las cursor rules que faltaban para showslink. https://t.co/mH392YnWr1 https://t.co/IVnvvvJM49

I use claude code with multi-session and multi-agent coding. I tried cmux, orca and other terminals. But I want to see claude sessions in one single page and switch to certain session easily. None of them felt right for me. So I built one for me. https://t.co/qzForQkXim

I also cancelled Cursor, as much as I love it and been using it for ages now. But I’ve relied more and more on Claude Code now, and in between rate limited windows i use Gemini CLI, so sadly had to cut some cost and Cursor was the one i used least https://t.co/DfQZ8iWmRz

I use both Claude Code and Codex, and I got tired of project knowledge staying with whichever agent I had been using. I moved the reviewed parts into the repo so both can work from the same context. https://t.co/Zi2l6j28IU

So I tried out Claude Code for 1 day. Overall I find the user experience quite smooth. My thoughts: - I used Claude Code to benchmark MiniMax M2.1 It does exactly what I say, but I felt I'm not exploring its capability enough - I still do a lot of hand holding to correct its cod…

Las skills fundamentales para trabajar con Claude Code y Codex son las de Matt Pocock. Si querés arrancar un proyecto o desarrollar uno mejor, pero no sabés por dónde empezar ni qué herramientas usar, esta es una buena forma de simplificarlo. Una skill es un archivo con instru…

88s para un task bastante básico, sin embargo, hubo un tremendo avance. El log de Claude Code me ayudo un montón a entender como funciona internamente. Me parece que voy a invertir un poco de tiempo en sacarle pasos. Menos `read` `search` y tareas básicas. El stack va a estar…

Claude Code me ayudó a solucionar el proyecto de reverse engineering que tenía 4 meses parado, voy a hacer el primero en sacar una app multiplataforma y open source para este dispositivo https://t.co/eICPmhYsZb

I’ve been using 3.7 with Claude Code terminal. I’m impressed how it will use tools like grep to find things and will create and run database migrations. I built something over the last 48 hours that would have taken a couple of weeks for sure. A new workflow I’ve been using rec…

After playing with Fable, GPT 5.6, and Opus 5 for the last couple of weeks, I think I nailed down my setup, for now... First, Claude Code as the harness. I’m doing more work in Codex these days. But still find the Claude Code harness better for long-running sessions because of…