¡Claude Code ahora también puede editar vídeos!
Esta SKILL es 100% gratis y de código abierto.
✓ Crea animaciones automáticamente
✓ Genera subtítulos con diferentes estilos
✓ Elimina silencios, errores y palabras de relleno
→ https://t.co/1SVpXY3BSG
13 years old.
making $20,000 a month.
working just 1-2 hours a day.
he didn’t work harder.
he built a system.
every morning, it looks at what’s trending.
it finds the best posts from the last 24 hours.
then it gives him 10 content ideas, ranked from best to worst.
he picks one.
one AI writes the post in his style.
another checks it for mistakes.
another posts it to 5 different platforms.
one idea.
five platforms.
almost no manual work.
after each post goes live, the system tracks what people liked, shared, and saved.
it uses that data to make the next posts better.
so the content keeps improving on its own.
the results add up.
5 million views in 2 weeks.
100,000 bookmarks in 2 months.
bookmarks matter.
people only save posts they want to come back to.
when lots of people save a post, platforms keep showing it to more people.
that’s why his posts keep getting views…
while he’s at school.
while he’s sleeping.
while he’s doing something else.
the biggest advantage today isn’t working more.
it’s building a system that works even when you don’t.
watch the video.
then read the full breakdown below.
Build an AI agent from first principles — without a framework hiding the loop.
AI Agents from Scratch is a local-first Python learning repo for developers who want to understand how agents actually work.
It helps you learn the mechanics by building one agent step by step from a single local LLM call, then adding structure, tools, memory, planning, evals, and telemetry.
Key features:
• 12-lesson path – moves from text-in/text-out to tools, loops, memory, planning, evals, and telemetry
• Local-first setup – uses a local GGUF model path, so you can learn without cloud APIs or API keys
• One evolving agent – the same https://t.co/O2lZ9lSfiA grows across lessons instead of hiding behavior behind a framework
• Runnable reference file – complete_example.py demonstrates all 12 lesson concepts in isolated functions
• Debugging + quality basics – evals and telemetry lessons cover regression tests, structured logging, latency, success rates, and traces
It’s open-source (MIT license).
Link in the reply 👇
🚨🚨 ESCÁNDALO: La población mundial recibió una "vacuna" para COVID diferente a la utilizada en los ensayos clínicos y completamente contaminada con ADN plasmídico, que contenía también un promotor del cancerígeno virus SV40...
"Para las pruebas clínicas antes de la aprobación, se utilizó una sustancia altamente pura. Era demasiado caro para una producción masiva... La población recibió una "vacuna" que se produjo con ayuda de la bacteria 'Escherichia coli'. El resultado es contaminaciones significativas con ADN bacteriano, y la consecuencia podría ser un aumento significativo del riesgo de cáncer"...
Dr. Helmut Sterz, el ex jefe de toxicólogos de Pfizer Europa.
Claude Code sin configurar es la mitad de lo que puede ser
Hasta que instalas claude-code-setup
El plugin oficial de Anthropic
Analiza tu proyecto y te dice exactamente que hooks, skills, MCP servers y subagentes activar
Ya no tienes que adivinar nada
Solo corres el comando y el setup se configura solo
/plugin install claude-code-setup@claude-plugins-official
Guarda esto para no perderlo 🔖
Loop Engineering is the next step after prompt engineering.
Most people still use Claude Code, Codex, Cursor, or Grok like a chat box:
Prompt.
Wait.
Copy.
Fix.
Prompt again.
This repo shows the next step:
You stop prompting the agent.
You design the loop that prompts the agent for you.
Inside:
→ Daily triage loops
→ PR babysitter loops
→ CI sweeper loops
→ Dependency sweeper loops
→ Changelog drafter loops
→ Post-merge cleanup loops
→ Issue triage loops
It also gives you CLIs to:
• Scaffold a loop
• Estimate token cost
• Audit if your repo is ready
• Add memory/state
• Add human handoff
• Add verification gates
• Run agents safely through GitHub Actions
The wild part is the shift in thinking.
Prompt engineering was about writing better instructions.
Loop engineering is about building a system where agents keep working, checking, fixing, and escalating without you babysitting every step.
This is what AI coding looks like when it stops being a chat session and starts becoming an operating system for software teams.
Repo: https://t.co/2USzC6KHUt
🤖 pentest-ai – AI Pentesting Tool That Verifies Every Finding
pentest-ai is an open-source AI-powered penetration testing framework that doesn't just detect vulnerabilities—it replays exploits to verify them before reporting. It combines AI agents, 200+ security tools, authenticated testing, exploit chaining, and machine-verified proof capsules to reduce false positives. With MCP support, CI integration, SARIF reports, and offline execution, it's built for modern security teams and researchers.
🔗 https://t.co/07LcoL7n0M
#CyberSecurity #Pentesting #AI #RedTeam #OpenSource
APPLE'S EULA SAYS YOU CANNOT INSTALL MACOS ON NON-APPLE HARDWARE.
A Brazilian developer named Gabriel Luchina built a one-line script that does exactly that. He runs a Portuguese Discord called Universo Hackintosh. He has been ignoring Apple's licensing terms in public for years.
The repo is called OSX-PROXMOX. It is sitting on GitHub at 7K stars.
Apple sells the cheapest Mac at $599. The cheapest Mac Studio at $1,999. They have spent decades building the legal and technical wall that says macOS only runs on their silicon.
This repo turns that wall into a single bash command:
/bin/bash -c "$(curl -fsSL https://t.co/XjnJpE6mok)"
Paste it into the Proxmox shell on any AMD or Intel box. You get a working macOS VM. Sequoia, Sonoma, Ventura, Monterey — your pick. With SIP enabled. With a properly signed OpenCore bootloader. On hardware Apple would refuse to acknowledge exists.
Here's the part Apple does not want trending:
→ It works on AMD, not just Intel
→ It supports eight macOS versions, from High Sierra to Sequoia
→ It installs in three steps from a fresh Proxmox install
→ It supports cloud deployment on VultR
→ It is maintained, with OpenCore 1.0.4 shipped April 2025
7K stars. 572 forks. 126 watchers. 12 releases. Active.
One honest note: the README explicitly says "for development, student, and testing purposes only." Apple's macOS EULA prohibits running macOS on non-Apple hardware. The legality depends on where you live and what you use it for. Read it.
Apple wants you to buy the hardware to use the software. One developer in Brazil has been quietly proving for years that the wall between them was always a license, not a chip.
Repo in the first comment.
TINYWALL es la mejor DEFENSA en tu WINDOWS, bloquea todo y tú decides #cortafuegos#gratis COMPARTE!
El tutorial completo está aquí: https://t.co/vYz5i1X51w
CLAUDE LANZÓ SU ACADEMIA GRATUITA DE IA 10 cursos oficiales con certificado incluido.
Desde cero hasta nivel avanzado. A tu ritmo.
Esto lo cambia todo.
Al final te explico cómo te ayudaría en tu desempeño dentro de los mercados.
Aquí tienes los 10 cursos (Guárdalos 🔖)