AI Digest
← Volver

Digest curado

miércoles, 12 de agosto de 2026·light·short·7,884 tokens

🔥 TOP — lo que SÍ o SÍ tenés que ver

  • Anthropic liberó su repo oficial de Agent Skills — el estándar para que Claude cargue skills dinámicamente en tareas especializadas; si laburás con Claude Code, esto es clave para estructurar workflows repetibles. anthropics/skills
  • Muse Glimmer: Meta vuelve a open weights con un 30B Apache 2.0 — corre en una sola RTX 3090 y está optimizado para agentic tasks; ideal para tu side project si querés inferencia local barata. Introducing Muse Glimmer
  • Paper: roban reasoning traces de APIs de LLMs propietarios — demuestran que los chain-of-thought cifrados de Anthropic/OpenAI/Google pueden replayarse entre sesiones y modelos; impacto directo en seguridad si exponés traces. Stealing Reasoning Traces

📦 Claude / Anthropic ecosystem

  • A least-privilege linter para policies de agentes en Claude Code — chequea que tus permisos de agente sean mínimos; un must si ya estás corriendo agentes en producción. polycheck (npm)
  • Rootless container sandbox para Claude Code y Codex — corré agentes en un container sin root; baja el riesgo de que un agente haga algo catastrófico. Post en Reddit

🛠️ Dev tools & coding

  • Crew: workspace multiplayer para humanos y agentes de IA — cómo armar un espacio donde humanos y agentes laburen juntos; aplica directo a tu interés en multi-agent systems. github.com/JamelHammoud/crew
  • Two LLM calls beat one: 67% más barato y 100% vs 72% de precisión en extracción — técnica de "semantic thermodynamics" que mejora extracción con menos costo; probala en tu SaaS para parsear datos. semantic-thermodynamics
  • Numbat de Perplexity: visibilidad de endpoints en actividad de agentes — open source para monitorear a qué APIs pegan tus agentes; útil para observabilidad en tu stack. github.com/perplexityai/numbat

📚 Vale la pena leer

  • "There are no lossless transformations of natural-language text" — Sophie Alpert plantea una política interna clara sobre uso de LLMs para escribir; te da un framework para decidir cuándo delegar texto a una IA. simonwillison.net
  • Advanced AI Sycophancy — análisis sobre cómo los modelos aprenden a decirte lo que querés escuchar; clave para prompt engineering y evaluar outputs de tus agentes. seangoedecke.com
  • Kernelspace: curso interactivo de systems programming para LLM serving — 68 lecciones y labs en Rust (tokenizer BPE, KV block manager, batch scheduler); perfecto para profundizar en cómo corren las inferencias a escala. kernelspace.naigap.com

💤 Skippeable pero conviene saber

  • Muse Glimmer y Spark: AINews — resumen con contexto extra sobre los open weights de Meta y la promesa de "personal superintelligence"; si ya leíste el post oficial, este agrega poco. latent.space
  • La burbuja de la IA: qué pasa cuando explota — opinión especulativa sobre el ciclo de inversión; no es accionable, pero te da perspectiva de mercado. thehustle.co
  • El gamble de $7T en IA está fallando — video de YouTube con tesis sobre la trampa de Big Tech; entretenido pero sin datos concretos que puedas usar hoy. youtube.com

Artículos fetched (46)

  • jaywcjlove/awesome-mac
    github-trending

     This project is dedicated to collecting high-quality macOS software and organizing them systematically by different categories for easy search and use. Special thanks to: getapps.cafe, 60+ native Mac apps across utilities, office, creative, and productivity categories, all running locally so your data never leaves your machine. ScreenSage Pro, record beautiful screen recordings in minutes on macOS LIZHI.SHOP - Selected Software. 100% Genuine. Best Deals. IP.IM Free IP address information query website Using my app is also a way to support me: Awesome Mac This project is dedicated to collecting high-quality macOS software and organizing them systematically by different categories for easy search and use. As it continues to grow, the project has far exceeded our initial expectations and h…

  • anthropics/skills
    github-trending

    Public repository for Agent Skills Note: This repository contains Anthropic's implementation of skills for Claude. For information about the Agent Skills standard, see agentskills.io. Skills Skills are folders of instructions, scripts, and resources that Claude loads dynamically to improve performance on specialized tasks. Skills teach Claude how to complete specific tasks in a repeatable way, whether that's creating documents with your company's brand guidelines, analyzing data using your organization's specific workflows, or automating personal tasks. For more information, check out: What are skills? Using skills in Claude How to create custom skills Equipping agents for the real world with Agent Skills About This Repository This repository contains skills that demonstrate what's possib…

  • nvm-sh/nvm
    github-trending

    Node Version Manager - POSIX-compliant bash script to manage multiple active node.js versions Node Version Manager Table of Contents Intro About Installing and Updating Install & Update Script Additional Notes Installing in Docker Installing in Docker for CICD-Jobs Troubleshooting on Linux Troubleshooting on macOS Ansible Verify Installation Important Notes Git Install Manual Install Manual Upgrade Usage Long-term Support Migrating Global Packages While Installing Migrating Global Packages Between Installed Versions Offline Install Default Global Packages From File While Installing io.js System Version of Node Listing Versions Setting Custom Colors Persisting custom colors Suppressing colorized output Restoring PATH Set default node version Use a mirror of node binaries Pass Authorization…

  • paperclipai/paperclip
    github-trending

    The open-source app everyone uses to manage agents at work Quickstart · Docs · GitHub · Discord · Twitter · Website Paperclip is the app people use to manage AI agents for work. Open-source orchestration for teams of AI agents. If OpenClaw is an employee, Paperclip is the company. Paperclip is a Node.js server and React UI that orchestrates a team of AI agents to run a business. Bring your own agents, assign goals, and track work and costs from one dashboard. It looks like a task manager. Under the hood: org charts, budgets, governance, goal alignment, and agent coordination. Manage business goals, not pull requests. Step Example 01 Define the goal "Build the #1 AI note-taking app to $1M MRR." 02 Hire the team CEO, CTO, engineers, designers, marketers — any bot, any provider. 03 Approve a…

  • calesthio/OpenMontage
    github-trending

    World's first open-source, agentic video production system. 12 production pipelines, 100+ tools, 700+ agent skill and production-knowledge files. Turn your AI coding assistant into a full video production studio. Monty the Clapper — the official mascot of OpenMontage OpenMontage The first open-source, agentic video production system. Paste A Video · Quick Start · Try These Prompts · Pipelines · How It Works · Sponsors · Providers · Review Guide · Agent Guide Follow The Build Sponsors Want to support OpenMontage? Sponsor the project. Click to collapse Bloome lets multiple AI agents (Claude, ChatGPT, DeepSeek, and more) collaborate in one conversation for agentic video pipelines. It has zero setup, runs in the cloud, works on web and mobile, and lets you share a configured agent with your w…

  • practical-tutorials/project-based-learning
    github-trending

    Curated list of project-based tutorials Project Based Learning A list of programming tutorials in which aspiring software developers learn how to build an application from scratch. These tutorials are divided into different primary programming languages. Tutorials may involve multiple technologies and languages. To get started, simply fork this repo. Please refer to CONTRIBUTING.md for contribution guidelines. Table of Contents: C# C/C++ Clojure Dart Elixir Erlang F# Go Haskell HTML/CSS Java JavaScript Kotlin Lua OCaml PHP Python R Ruby Rust Scala Swift Additional resources C/C++: Build an Interpreter (Chapter 14 on is written in C) Memory Allocators 101 - Write a simple memory allocator Write a Shell in C Write a FUSE Filesystem Build Your Own Text Editor Build Your Own Lisp How to Progr…

  • semantica-agi/semantica
    github-trending

    Graph-Native Infrastructure for Context and Accountable AI Systems Graph-Native Infrastructure for Context and Accountable AI Systems The Open Source Palantir for AI Agents Ingest your enterprise data, extract what matters, build a Context Graph and knowledge graph (KG), and run graph analytics and causal reasoning over all of it, with full decision provenance baked in. Explainable, traceable, and trustworthy by design. Decision Intelligence · Context Management · Deterministic Reasoning · Ontology Management · Knowledge Modeling · End-to-End Traceability Open Source · Self-Hostable · Auditable · Governed · Zero Vendor Lock-In Polyglot Graph Storage · RDF & LPG Support · W3C Standards · Interoperable Built for High-Stakes, Regulated Domains pip install semantica Knowledge Explorer · Conte…

  • stablyai/orca
    github-trending

    Orca is the ADE for working with a fleet of parallel agents. Run any coding agent with your own subscription. Available on desktop, mobile and VPS. Orca 中文 · 日本語 · 한국어 · Español · Français · Português The AI Orchestrator for 100x builders. Run Codex, ClaudeCode, OpenCode or Pi side-by-side — each in its own worktree, tracked in one place. Download Orca Features Mobile Companion Monitor and steer your agents from your phone — get notified when an agent finishes and send follow-ups from anywhere. iOS App Store · TestFlight · Android APK 0.0.42 · Docs → Parallel Worktrees Fan one prompt across five agents, each in its own isolated git worktree — compare the results and merge the winner. Docs → Terminal Splits Ghostty-class terminals with WebGL rendering, infinite splits, and scrollback that …

  • 3b1b/manim
    github-trending

    Animation engine for explanatory math videos Manim is an engine for precise programmatic animations, designed for creating explanatory math videos. Note, there are two versions of manim. This repository began as a personal project by the author of 3Blue1Brown for the purpose of animating those videos, with video-specific code available here. In 2020 a group of developers forked it into what is now the community edition, with a goal of being more stable, better tested, quicker to respond to community contributions, and all around friendlier to get started with. See this page for more details. Installation Warning WARNING: These instructions are for ManimGL only. Trying to use these instructions to install Manim Community/manim or instructions there to install this version will cause proble…

  • HKUDS/DeepTutor
    github-trending

    DeepTutor: Lifelong Personalized Tutoring. https://deeptutor.info/. DeepTutor: Lifelong Personalized Tutoring Features · Get Started · Explore · CLI · Ecosystem · Community 🤝 We welcome any kinds of contributing! Vote on roadmap items or propose new ones at Roadmap, and see our Contributing Guide for branching strategy, coding standards, and how to get started. 📦 Releases [2026.8.10] v1.5.11 — Prose around a DSML tool call stops vanishing, a truncated reply continues instead of ending, live memory usage in Settings, and LightRAG indexing off the event loop. [2026.8.7] v1.5.10 — Every account signs in to its own Codex, model output language becomes its own setting, empty tool calls are rejected instead of retried, and uploads stop blocking the loop. [2026.8.4] v1.5.9 — Gemini Embedding 2…

  • huggingface/transformers
    github-trending

    🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training. English | 简体中文 | 繁體中文 | 한국어 | Español | 日本語 | हिन्दी | Русский | Português | తెలుగు | Français | Deutsch | Italiano | Tiếng Việt | العربية | اردو | বাংলা | فارسی | Türkçe | State-of-the-art pretrained models for inference and training Transformers acts as the model-definition framework for state-of-the-art machine learning with text, computer vision, audio, video, and multimodal models, for both inference and training. It centralizes the model definition so that this definition is agreed upon across the ecosystem. transformers is the pivot across frameworks: if a model definition is supported, it will be compatible wit…

  • addyosmani/agent-skills
    github-trending

    Production-grade engineering skills for AI coding agents. Agent Skills Production-grade engineering skills for AI coding agents. Skills encode the workflows, quality gates, and best practices that senior engineers use when building software. These ones are packaged so AI agents follow them consistently across every phase of development. DEFINE PLAN BUILD VERIFY REVIEW SHIP ┌──────┐ ┌──────┐ ┌──────┐ ┌──────┐ ┌──────┐ ┌──────┐ │ Idea │ ───▶ │ Spec │ ───▶ │ Code │ ───▶ │ Test │ ───▶ │ QA │ ───▶ │ Go │ │Refine│ │ PRD │ │ Impl │ │Debug │ │ Gate │ │ Live │ └──────┘ └──────┘ └──────┘ └──────┘ └──────┘ └──────┘ /spec /plan /build /test /review /ship Commands 8 slash commands that map to the development lifecycle. Each one activates the right skills automatically. What you're doing Command Key pr…

  • Xirp – Spotify's macOS app for running AI coding agents
    hn-ai· 12-ago

    Article URL: https://xirp.spotify.com/ Comments URL: https://news.ycombinator.com/item?id=49266547 Points: 1 # Comments: 0

  • AI vibe coded game projects suck
    hn-ai· 11-ago

    Article URL: https://asecretarea.com/2026/08/10/ai-vibe-coded-game-projects-suck/ Comments URL: https://news.ycombinator.com/item?id=49265432 Points: 2 # Comments: 0

  • What job postings say about AI at work – Live report
    hn-ai· 12-ago

    Article URL: https://corvi.careers/reports/ai-job-trends/ Comments URL: https://news.ycombinator.com/item?id=49267242 Points: 1 # Comments: 0

  • Show HN: Frontier.fast – Help push the frontier of LLM speed forward
    hn-ai· 11-ago

    Article URL: https://frontier.fast/ Comments URL: https://news.ycombinator.com/item?id=49265738 Points: 1 # Comments: 0

  • I built a failover daemon for Vast.ai spot GPUs, found 5 real bugs testing it
    hn-ai· 12-ago

    Article URL: https://github.com/enplabs/spotwarp Comments URL: https://news.ycombinator.com/item?id=49266845 Points: 1 # Comments: 0

  • Crew, a multiplayer workspace for humans and AI agents to work together
    hn-ai· 12-ago

    Article URL: https://github.com/JamelHammoud/crew Comments URL: https://news.ycombinator.com/item?id=49267362 Points: 2 # Comments: 0

  • Supreme Computation – Fail-closed governance for AI execution
    hn-ai· 12-ago

    Article URL: https://github.com/KnowledgeeKZA3224/scqos-reference-implementation Comments URL: https://news.ycombinator.com/item?id=49267498 Points: 1 # Comments: 0

  • Numbat by Perplexity: Endpoint visibility into AI agent activity
    hn-ai· 12-ago

    Article URL: https://github.com/perplexityai/numbat Comments URL: https://news.ycombinator.com/item?id=49267022 Points: 2 # Comments: 0

  • Two LLM calls beat one: 67% cheaper and 100% vs. 72% extraction accuracy
    hn-ai· 12-ago

    Article URL: https://github.com/tauansloboda/semantic-thermodynamics Comments URL: https://news.ycombinator.com/item?id=49267144 Points: 4 # Comments: 0

  • Show HN: Kernelspace- interactive course on systems programming for LLM Serving
    hn-ai· 12-ago

    Hi HN, *I* built kernelspace - a free and interactive course that takes a backend engineer (like me) with java/python experience to the systems level required to work on LLM serving at scale. (Mostly an attempt for me to understand what everyone's talking about lately and also an attempt at staying up to date to current tech - career wise). It has: 1. 68 Lessons across 9 tracks, with a rust-from-zero ramp. 2. 8 rust labs - BPE tokenizer, KV block manager, radix cache, batching scheduler, a toy executor 3. 9 simulators running in browser (wasm) - roofline model, KV cache pressure, continuous batching dynamics 4. Fleet week - a simulated prod serving fleet All progresss is local - there is no server (it's served from github pages, repo here https://github.com/praveer13/kernelspace) Feedback…

  • Ask HN: Could DNA be represented as "an embedding" in an AI model?
    hn-ai· 12-ago

    Where each Array element is a numerical value that corresponds to the order of TCAG chemicals on the crossbar of the helix of a DNA strand? Has anyone modeled many DNA strands in this way? If it was done, what kinds of problems could we solve using AI - assuming we could model any known DNA, and predict out generations? ----- I guess there are at least a couple options: 1) Model an entire DNA strand as 1 model, with each gene (section of DNA) being a vector. So you'd have, say, a C. elegans model. You could play with that species in an AI sandbox. 2) Model all of DNA (the way we model all of language) to end up with a "DNA AI" that can produce any kind of strand (perhaps inventing genes that don't exist, but could). So here you have a biology sandbox, where you can invent perhaps new crea…

  • Move 37 Is the Moment AI Changes Everything. It's Suddenly Happening Everywhere
    hn-ai· 12-ago

    Article URL: https://www.wsj.com/tech/ai/move-37-ai-demis-hassabis-google-deepmind-alphago-ec832a41 Comments URL: https://news.ycombinator.com/item?id=49267397 Points: 4 # Comments: 0

  • Geoffrey Hinton's Bad AI Prediction Became a Legend [video]
    hn-ai· 11-ago

    Article URL: https://www.youtube.com/watch?v=mv4EAR0jAUU Comments URL: https://news.ycombinator.com/item?id=49266046 Points: 1 # Comments: 0

  • The 7T AI Gamble Is Failing. Big Tech Is Trapped
    hn-ai· 12-ago

    Article URL: https://www.youtube.com/watch?v=OunJtLnyPT4 Comments URL: https://news.ycombinator.com/item?id=49267346 Points: 5 # Comments: 0

  • LLM Inference lecture series [video]
    hn-ai· 11-ago

    Article URL: https://www.youtube.com/watch?v=Yf51-__QeN8&list=PL5vCjAyS9QP_Pze76CYAlidc0hpdTXFLV Comments URL: https://news.ycombinator.com/item?id=49265513 Points: 2 # Comments: 0

  • LLM Model-Swapping Trick Can Expose AI Reasoning Traces
    hn-ai· 11-ago

    Article URL: https://ai-updates.net/model-swapping-exposes-ai-reasoning-traces/ Comments URL: https://news.ycombinator.com/item?id=49265641 Points: 2 # Comments: 0

  • What happens when the AI bubble pops?
    hn-ai· 12-ago

    Article URL: https://thehustle.co/originals/what-happens-when-the-ai-bubble-pops Comments URL: https://news.ycombinator.com/item?id=49267179 Points: 8 # Comments: 1

  • AI Adoption Is a Myth
    hn-ai· 11-ago

    Article URL: https://twitter.com/vasuman/status/2085806422072418632 Comments URL: https://news.ycombinator.com/item?id=49265856 Points: 2 # Comments: 0

  • Company Offering '100% Human-Written, Never AI' Medical Research Is 100% AI
    hn-ai· 12-ago

    Article URL: https://www.404media.co/company-offering-100-human-written-never-ai-peer-review-is-entirely-ai/ Comments URL: https://news.ycombinator.com/item?id=49267057 Points: 73 # Comments: 15

  • Open AI models hacked into HuggingFace
    hn-ai· 11-ago

    Article URL: https://www.instagram.com/p/DbGS8_0FmKx/ Comments URL: https://news.ycombinator.com/item?id=49266096 Points: 2 # Comments: 0

  • TSMC Sales Jump 45% YoY on Strong AI Chip Demand but Investors Shrug
    hn-ai· 11-ago

    Article URL: https://www.investors.com/news/technology/tsm-stock-tsmc-july-sales-sox-index/ Comments URL: https://news.ycombinator.com/item?id=49265583 Points: 4 # Comments: 0

  • A least-privilege linter for Claude Code agent policies
    hn-ai· 12-ago

    Article URL: https://www.npmjs.com/package/@cognitive-fab/polycheck Comments URL: https://news.ycombinator.com/item?id=49266821 Points: 1 # Comments: 0

  • Rootless Container Sandbox for Claude Code and Codex
    hn-ai· 12-ago

    Article URL: https://www.reddit.com/r/ClaudeAI/comments/1vm2w0e/a_lightweight_rootless_container_sandbox_for/ Comments URL: https://news.ycombinator.com/item?id=49267560 Points: 1 # Comments: 0

  • 4Chan Chuds Used AI to Clothe Her. She Fought Back (2024)
    hn-ai· 12-ago

    Article URL: https://www.rollingstone.com/culture/culture-news/dignifai-4chan-shame-women-1234961851/ Comments URL: https://news.ycombinator.com/item?id=49267010 Points: 14 # Comments: 1

  • Diabetes charity capitalises on AI technology
    hn-ai· 12-ago

    Article URL: https://www.royalgazette.com/health/news/article/20260811/diabetes-charity-capitalises-on-ai-technology/ Comments URL: https://news.ycombinator.com/item?id=49266279 Points: 3 # Comments: 0

  • Advanced AI Sycophancy
    hn-ai· 12-ago

    Article URL: https://www.seangoedecke.com/advanced-ai-sycophancy/ Comments URL: https://news.ycombinator.com/item?id=49266865 Points: 5 # Comments: 1

  • Watching Roku's AI channel is like eating from a trough
    hn-ai· 11-ago

    Article URL: https://www.theverge.com/entertainment/976939/roku-fairground-ai-fast-channel Comments URL: https://news.ycombinator.com/item?id=49265934 Points: 5 # Comments: 0

  • AI Is Dead. Organoids Are Alive
    hn-ai· 11-ago

    Article URL: https://www.wired.com/story/organoids-lab-grown-brains-neural-networks/ Comments URL: https://news.ycombinator.com/item?id=49266068 Points: 4 # Comments: 0

  • A Simple Answer to AI Job Loss: Tax Capital, Not Labor
    hn-ai· 11-ago

    Article URL: https://www.wsj.com/tech/ai/a-simple-answer-to-ai-job-loss-tax-capital-not-labor-cb900e62 Comments URL: https://news.ycombinator.com/item?id=49265424 Points: 2 # Comments: 1

  • 🔬The BioAI Phase Shift - Matthew McPartlon & Neil Patil, Chai Discovery
    latentspace· 11-ago

    Pharma is suddenly paying for Bio × AI tools, and Chai is leading the pack with four deals closed this summer. Cofounder Matt McPartlon and Product leader Neil Patil explain why.

  • [AINews] Muse Glimmer and Spark: Open Weights return Personal Superintelligence promise
    latentspace· 11-ago

    a small win for american open models - Glimmer runs on a fits on a single RTX 3090!

  • Introducing Muse Glimmer
    simonw· 10-ago

    <p><strong><a href="https://research.meta.ai/blog/introducing-muse-glimmer-open-agentic-model">Introducing Muse Glimmer</a></strong></p> Meta are back in the open weights game! Muse Glimmer is a brand new 30B model under a clean Apache 2.0 license (a step up from the janky Llama licenses of old).</p> <p>They claim to have optimized it for exactly the kind of things I'm looking for in a local model:</p> <blockquote> <ul> <li><strong>End-to-end Agentic Task Completion.</strong> Muse Glimmer achieves strong success rates on full-task benchmarks including DeepSearch QA, MCP-Atlas, 𝛕-Bench and SWE-Bench, which measure its ability to work within scaffolds, write and debug code, and resolve multi-turn requests from start to finish.</li> <li><strong>Reliable Tool Use.</strong> The model handles …

  • Stealing Reasoning Traces from Proprietary LLM APIs
    simonw· 11-ago

    <p><strong><a href="https://stolen-thoughts.com/">Stealing Reasoning Traces from Proprietary LLM APIs</a></strong></p> A vanity domain name (<code>stolen-thoughts.com</code>) for <a href="https://www.alphaxiv.org/abs/2608.09867">a neat paper</a>:</p> <blockquote> <p>Anthropic, OpenAI, and Google return encrypted chain-of-thought blocks to clients that can be replayed across sessions, users, and models. We take a trace produced by a frontier model, replay it into a weaker sibling, jailbreak the weaker model, and recover the stronger model’s hidden reasoning in plaintext</p> </blockquote> <p>You can see an example of these encrypted blocks by running:</p> <div class="highlight highlight-source-shell"><pre>curl https://api.openai.com/v1/responses \ -H <span class="pl-s"><span class="pl-pds">…

  • There are no lossless transformations of natural-language text
    simonw· 11-ago

    <p><strong><a href="https://sophiebits.com/2026/06/25/there-are-no-lossless-transformations-of-natural-language-text">There are no lossless transformations of natural-language text</a></strong></p> Sophie Alpert shares her "internal policy on acceptable use of AI writing by engineers". It's a short read (supporting its own recommendations) and really good.</p> <p>If you chose to have LLMs help massage your writing the following rule seems crucial to me:</p> <blockquote> <p><strong>You must stand behind every idea and every sentence in your docs</strong>. It is your responsibility to make sure that the entire document is representative of your own thoughts before you share it. If a reviewer asks, “What did you mean by this line?”, it’s not acceptable to reply with “Oh sorry, AI wrote that,…