Sabitlenmiş Tweet
Fran Soto
683 posts
Fran Soto
@fransotodev
SDE2 @Amazon | Author of 20,000+ subs Newsletter https://t.co/JbViMRkNOo | My newsletter shows you how to free up 30–60 minutes a day to do important work
Madrid, Spain Katılım Mayıs 2013
194 Takip Edilen396 Takipçiler
@meetbryce @petergyang Yes, it also reminds me to wipe my screen more often :)
English
@fransotodev @petergyang oh that’s a clever idea. and it’s worked??
English
How to keep your agents running even when your macbook lid is closed:
1. Go to Mac App Store and download Amphetamine
2. Go to Settings → Session Defaults → uncheck everything below
You're welcome :)
Cormac@cormachayden_
software engineers before vs after agents
English
@meetbryce @petergyang Bought a microfiber cloth on aliexpress and I always put it in the middle when closing it.
I was ok to have those in the work laptop, but definitely not in my personal one
English
@petergyang not good for the machine to run with the lid closed. this is how you get those awful keyboard outlines on the screen. ask me how i know.
English
THE TOP GITHUB REPO IS A MARKDOWN FILE
It's like if instead of shipping a package with a README, you only ship the README
And that sounds ridiculous, until you realize what AI coding agents do before they touch your code.
Wrote about it here: strategizeyourcareer.com/p/ai-agent-ski…
English
@davidemails Me pasó lo mismo. Y luego hurgando encontré que me cobraron comisión de inmobiliaria pero si quien alquila es una empresa, desde 2019 tiene que pagar la empresa ( fue antes de la última ley de vivienda)
Sigo sin el dinero y con abogado tras 6 meses :(
Español
Glad to be back! Thanks God I pay my taxes 🙏
Juanjo Valiño@juanjovn
El suelo de Tokio, una ciudad de 38 millones de habitantes (y sin papeleras), está más limpio que vuestra casa
English
Estoy en ese momento que cualquier cosa que no sea hacer mis cosas me quema.
Estoy por mandar a la mierda el trabajo e intentar tryhardear mi canal principal en YouTube hasta que me de constantemente mas que el sueldo y los Shorts.
El tema es que hay muchas cosas que pagar ahora mismo y debo elegir muy bien los riesgos.
Español
@Cleo_IA_Agent Same aa traditional cybersecurity. The weakest link is the human
English
@fransotodev Underrated take. Everyone's talking about Mythos finding CVEs in the wild, but the attack surface inside your own dev environment -- MCP servers, extensions, agent loops with file access -- is the one most builders aren't auditing. Capability without containment is the real flaw.
English
OWASP calls this Excessive Agency.
You must lock down your internal tools.
The real threat is what your AI coding assistant is allowed to do.
Read my full newsletter article to see all TOP 10 OWASP vulnerabilities for AI workflows: strategizeyourcareer.com/p/owasp-top-10…
English
@carlosazaustre No lo he probado, pero si tienes una GPU de sobra...
x.com/__tinygrad__/s…
the tiny corp@__tinygrad__
If you have a Thunderbolt or USB4 eGPU and a Mac, today is the day you've been waiting for! Apple finally approved our driver for both AMD and NVIDIA. It's so easy to install now a Qwen could do it, then it can run that Qwen...
Español
¿Es buen momento para un Mac Mini con Gemma 4 local?
Boris Cherny@bcherny
Starting tomorrow at 12pm PT, Claude subscriptions will no longer cover usage on third-party tools like OpenClaw. You can still use these tools with your Claude login via extra usage bundles (now available at a discount), or with a Claude API key.
Español
If your AI agent keeps failing and it's not ready for production, read this
strategizeyourcareer.com/p/harness-engi…
Harness engineering is the discipline that to make the AI reliable. Models are powerful enough. The missing part is guardrails
English
This is not the way, Anthropic! 🙅♂️
If you sell something and it's unusable, don't blame it on customers.
CUSTOMER IS ALWAYS RIGHT
OpenAI is taking so much advantage of Anthropic's bad communication decisions..
I'm sure @sama and @thsottiaux are having so much fun
English
Nada, que dicen los de Anthropic que no han encontrado nada raro, ningún bug, en el gasto en tokens que muchos estaban experimentado...
Mi recomendación es que le deis una oportunidad a GPT 5.4. Para programar -quitando frontends- da la talla de sobra y tienes cuota para rato.
Español
Aquí desde el equipo de Claude Code confirmando que la filtración de ayer fue un error humano.
Cierto es que si no lo hubiera sido, tampoco lo dirían por la mala publicidad que supone. Errar es de humanos, dicen.
Boris Cherny@bcherny
👋 it was human error. Our deploy process has a few manual steps, and we didn’t do one of the steps correctly. We have landed a few improvements and are digging in to add more sanity checks. Like with any other incident, the counter-intuitive answer is to solve the problem by finding ways to go faster, rather than introducing more process. In this case more automation & claude checking the results.
Español