Brayan Angulo @Briamcode

RFE/RL takes apart a Russian Geran-4 drone to uncover what makes the jet-powered unmanned aerial vehicle (UAV) faster, harder to intercept, and capable of operating deep inside Ukraine -- from its turbojet engine and advanced camera to foreign-made communications and navigation components. @cxemu @radiosvoboda

Turn any open LLM into a Jev-style model: Researchers just open-sourced AnyJev, a Python library that extracts typed choices, yes/no judgments, scores, and probabilities from causal LLMs. Beyond its raw baseline, AnyJev offers three levels depending on how much labeled data and calibration you need: - L0 requires no labels. It rotates the answer options to reduce position bias, but its probabilities remain uncalibrated. - L1 uses 100-500 labeled examples per question to calibrate those probabilities through temperature scaling. It improves confidence estimates without changing the answer ranking. - L2 uses 100-300 labels to fit a closed-form head on an intermediate hidden state. It leaves model weights untouched and can stop inference before the full forward pass completes. The main constraint is model access. AnyJev requires access to logits or hidden states, so it is built for open models. L2 heads are also specific to a model and question. It is compatible with several open causal LLMs. Repo: github.com/nokia-applied-… (don't forget to star it ⭐ ) If you want to dive deeper, I also wrote about a similar mechanism in a hands-on guide. It covers building a Jev-style decision path with open models, entirely locally. Read it below.

Avi Chawla @_avichawla

6 days ago

x.com/i/article/2101…

61 437 3K 539K 5K

this is f*cking gold. How to build your first AI agent with Jev. This is everything you need to know to be ahead of the most people. Once Jev is in your loop, your agent picks its own next action, rates its own outputs, and decides when the goal is met. it runs without you watching it. Jev is a decision model built specifically for this. it reads your system state and returns a typed answer with confidence: Choice - which agent or action should happen next? Score - is this result good enough to keep? Noul - is the goal met? the architecture: LLM thinks. Jev decides. Agent executes. 10 minutes to get it running. Hosted on Vercel and Cloudflare via Typesafe. Waitlist access, people report getting in within a day. The guide in the image maps the full setup, from architecture to your first decision call. Wrote the full breakdown on agents, loops, and graphs below.

Mahax @Mahaximus_

2 months ago

x.com/i/article/2084…

47 354 2K 4.9M 7K

Hardware voltage glitching with success in seconds: AES key recovery, authentication bypass, and more! 📟⚡🤖⏱️🏆 More details on: LinkedIn: lnkd.in/p/dUuwS5X9 Substack: it4sec.substack.com/p/hardware-vol… Telegram: t.me/it4sec/54

Exploit Released for Unpatched Ubuntu Linux Flaw Enabling Host-Root Container Escape thehackernews.com/2026/09/exploi…

El software del control deck solo existía para Windows. Le hice ingeniería inversa a la DLL, leí el tráfico USB y armé mi propio daemon en Python con libusb. Código abierto. #linux #reverseengineering #opensource

🚨 CRITICAL NEXT.JS RCE DISCLOSED — ATTACKER-CONTROLLED SVG DATA CAN LEAD TO SERVER CODE EXECUTION Vercel has disclosed a critical remote-code-execution vulnerability affecting the Node.js implementation of ImageResponse in Next.js. • CVE-2026-94545 / GHSA-vcvr-r3jv-pc5j • CVSS 9.5 Critical • Affects Next.js >=16.2.0 and <16.3.6 • Exploitation is network-accessible and requires no authentication or user interaction, but requires an affected application configuration • Applications are vulnerable when attacker-controlled values reach SVG content, attributes or styles rendered through next/og ImageResponse • Successful exploitation can lead to remote code execution on the Node.js server • The Edge implementation of ImageResponse is NOT affected • Applications that never place attacker-controlled data into generated SVG content, attributes or styles are also not affected • Fixed in Next.js 16.3.6 • Vercel recommends removing untrusted SVG input from Node.js ImageResponse as a workaround where immediate upgrading is impossible • No confirmed in-the-wild exploitation has been identified at this time ⚠️ Analyst Note: The interesting attack surface is dynamic image generation. Open Graph/social-preview images often incorporate URL parameters, usernames, titles or other externally supplied content — exactly the type of data that can become attacker-controlled. Internet-facing Next.js applications using server-side ImageResponse should therefore review not only their framework version but whether untrusted values flow into SVG generation. Original Vercel / Next.js security advisory: github.com/vercel/next.js… #NextJS #Vercel #CVE202694545 #RCE #WebSecurity #Vulnerability #ThreatIntel #DDW #DarkWeb

Build agentic workflows completely offline. The Antigravity SDK now supports local execution with Gemma 4 and LiteRT. Run agents entirely on your local machine with: 💵 Zero token costs 🔒 Total data privacy 🔌 Offline reliability Bonus feature: Support for OpenAI-compatible endpoints. Use Ollama, llama.cpp, vLLM and more to serve Gemma 💪 Get started: pip install google-antigravity litert-lm Read the details: developers.googleblog.com/introducing-su…

this is f*cking gold How to build your first AI agent (Full guide) this is basically everything you need to know about building with Jev in the right hands, this changes everything:

Khairallah AL-Awady @eng_khairallah1

3 days ago

Jev is the missing brain in every AI agent It decides what happens next in milliseconds, at almost zero cost Set it up right and you cut most of your agent bill without touching quality In this article, I show you how x.com/i/article/2102…

33 42 184 231K 370

BurpSuite is a must-know tool for hackers, pentesters, and bug bounty hunters. In this tutorial, we’ll dive into fuzzing—injecting random data to uncover vulnerabilities like SQL injection & XXE. hackers-arise.com/web-app-hackin…

Data recovery is mostly detective work, not magic, just a lot of patience and a steady hand. Also a great reminder that backups should exist before your panic does.

internet archiva @internetarchiva

6 months ago

This is how engineers bring back “lost data”

2 45 217 1.1M 109

🚨 THIS IS WHAT JEV LOOKS LIKE IN THE REAL WORLD. Browser Use just built a Jev-powered browser agent that can complete real tasks in seconds. This is Jev Ultrafast. Give it one goal. It decides what to do + which element to use from the current page. → Dynamic action space built from the live DOM → One TypeSafe request per decision cycle → Click, type, select, scroll, wait, done → Small LLM only generates text when needed → No screenshots required in the default loop → Built-in target validation before browser actions And the demo is crazy: Google Flights → Zürich → London → 7.1 seconds. The repo reports a median task time improvement from 9.450s → 7.092s, while browser protocol calls dropped from 1,092 → 101 in its six-run comparison. Basically: Jev decides → Browser Use executes → small LLM types → task finishes. This is a much more interesting way to think about browser agents: Don’t make the AI generate everything. Make it decide what action comes next. 🔥 #Jev #BrowserUse #AIAgents #AI #Automation #LLM #TypeSafe #BrowserAutomation #OpenSource #Developers

Vikas gupta @vicky_grok

5 days ago

JEV: THE AI MODEL THAT DOESN’T NEED TO GENERATE TEXT. Most AI generates text token by token. Jev takes a different approach. No text-first generation. A completely different way to build AI. x.com/i/article/2101…

2 0 11 29K 12

THE MOST POPULAR JEV REPO ON GITHUB IS 21 FILES OF PYTHON ~17,000 stars. 1,074 forks. six days old. i swept all 1,632 repos in this ecosystem - 6.3 GB, all of it younger than a week to find one with actual architecture behind it. then i rendered every line of it. 96,167 source lines -> STARS MEASURE THE DEMO, NOT THE CODE. the top two repos in this ecosystem have 23,000 stars and 35 code files between them. the one worth rendering has 70 stars, 1,314 files, and an owner who has been shipping since 2009 with 673 repos behind him. 32 of the 1,632 are awesome-lists pointing at each other. check the ACCOUNT, not the badge. would you let something that cannot explain itself decide which of your rules fires

barnyx @me_barnyx

5 days ago

x.com/i/article/2101…

22 28 203 233K 423

Jev has been blowing up lately. If you've got the Jev API but don't know how to play around with it yet, you can just copy this checklist. 1. jev-ultrafast A high-speed browser Agent built with Browser Use. Jev only judges "what to do, which element to click" at each step, and only calls the small model when typing is needed. Searching for a flight on Google Flights takes about 7 seconds. github.com/browser-use/je… 2. fast-jev-compaction Context compression for Claude Code. Before each tool call, have Jev judge if there's anything still useful; delete the useless stuff, and keep the original text without rewriting it. github.com/tamaratran/fas… 3. json-render Vercel Labs' generative UI framework. In experiments, Jev doesn't write JSON token by token; it just handles selecting components, properties, and layouts. github.com/vercel-labs/js… 4. typesafe-mcp Best for people who just got the API. Plug Jev into Claude Code, Claude Desktop, Codex, and Pi, and do Choice / Score / Noul anytime. github.com/itsmostafa/typ… 5. jev-mcp Ready-made Agent judgment toolkit: fact-checking, content screening, semantic ranking, classification, and information extraction. github.com/jkudish/jev-mcp 6. SemDecide Turn Jev into a command-line tool. Directly classify, score, and filter in the Shell—great for hooking up to crawlers, CI, and data pipelines. github.com/sharziki/semde… 7. jev-codex-router First have Jev judge how hard this round of programming tasks is, then decide the model tier, reasoning depth, and speed mode. github.com/0xNatoshi/jev-… 8. Winnow Context garbage collection for Claude Code. When Read / Bash / Grep spits out a ton of stuff, Jev first judges which parts are really relevant to the current task. github.com/GhalebDweikat/… 9. jev-review Before code review, run it through Jev first to pick out high-risk changes, then hand them off to a pricier big model or a human. Comes with a local dashboard. github.com/devagrawal09/j… 10. Blink Use Jev as a code repository navigator. At each directory level, judge which files are most relevant to the current issue, then keep digging down. github.com/ellipsis-dev/b… Copy these complete Jev blueprints - then read full Jev setup below ↓ ↓

darkzodchi @zodchiii

a week ago

x.com/i/article/2101…

17 91 575 1.4M 2K

🔍 OWASP Nettacker: Automated Security Scanning A quick 15-second look at Nettacker, an open-source framework for automated reconnaissance, information gathering, and vulnerability assessment. 💻🔐 � GitHub 🛡️ Authorized testing only. #OWASPNettacker #CyberSecurity #InfoSec

Jev Founder, Diogo Amogo, just released 12-page PDF on building a Jev Harness for coding agents this is a 10-step blueprint on how to make your coding agents 200× faster and 400× cheaper: step 1 → meet Jev: the LLM writes, the harness executes, Jev decides what each turn sees, where it routes and whether it runs step 2 → ask the question that breaks every agent: how would you design one if LLMs had no KV cache? step 3 → stop routing blind: Opus → Sonnet → Opus costs 6.19 vs 4.15 for pure Opus, because handing back reprocesses the whole context step 4 → follow the tokens: reading and searching take 56.2% of tool turns and 46.5% of tokens. Writing code is under 10% step 5 → score every chunk per query: hide, short summary, long summary or full. Compress after the question, not before step 6 → disclose tools in tiers: one-line snippets for 100s of tools, schema on demand, docs for one-off queries. Batteries stop costing context step 7 → load instructions by condition: touching *.tsx loads the style guide, billing/ loads its gotchas file, and compaction can't erase either step 8 → route by trust, not just difficulty: secrets and infra stay on first-party frontier models, public docs go to the cheapest one step 9 → share one retrieval pass: cross-model review, eval generation, ELI5 explainers and live progress pages all run read-only in the background step 10 → gate every command: programmable allow / ask / deny policies that read a script before it runs, not just its name Send this PDF and the article below to your Claude Code or Codex instance and start shipping 200× faster.

Movez @0xMovez

a week ago

x.com/i/article/2100…

36 117 883 1.1M 2K

I have released gem-x.cpp on GitHub! This produces a 3D skeleton from video, including live video, which is what interests me. NVIDIA showed the original GEM-X teleoperating humanoids, so the next thing I want to do is get that working in simulation all in GGML/C++.

Tutorial completo de Jev Qué es, cómo puedes crear con él y qué nuevas aplicaciones puede desbloquear → 0:00 Introducción → 0:34 Jev explicado → 4:06 Configuración de la API → 5:59 Demo 1: Navegador controlado por voz → 11:33 Demo 2: Memoria de IA → 17:27 Demo 3: Predictor de YouTube

JEV BOT ON GOD-MODE there is no catch and that scares me Elon Musk posted an idea that hit 24 million views: "the edge disappears the second everyone can see it" that's basically what Jev Bot does now - I gave it $50 and told it to earn its keep or get shut off $50 → $8,310 in 22 hours still running, still compounding - nobody's touched it since it started, it just sits on its own machine in the cloud while the markets move somewhere else every 5 minutes it: > scans every open market before the crowd catches on > checks how far the real signal has pulled ahead of what the posted odds still assume > only steps in once that gap clears a set threshold > sizes the position to how wide the gap is > executes on its own, no sign-off from me > pays its own hosting bill from whatever it made that week miss a payment and it's dead - no warning, no grace period after dropping to $7 on hour two, it stopped touching anything where the signal and the odds still disagreed - waited for the move to be undeniable instead of guessing setup was one evening: build it, run a single trade in front of it, schedule it, connect a wallet no VPS. no API keys. no code to maintain some of these have been running for weeks without anyone touching them again - you just don't hear about it until someone finally posts bookmark this before you're the one still watching

bl888m @bl888m_eth

a week ago

x.com/i/article/2100…

4 16 137 541K 442

Adding Jev to an AI agent does not automatically make the loop faster the speed comes from finding decisions the agent makes over and over, then routing them deliberately. formula: Faster Agent Loop = Trace + Jev Routing + Model Fallback + Verification 10% -> capture the loop. log model calls, tool actions, retries, latency and cost so you know where time goes. 24% -> group repeated decisions. identify choices like “use a tool?”, “retry?” and “is this done?”, then define the correct outcome for each. 56% -> give Jev the relevant state and clear criteria. let it route routine decisions without sending the entire conversation through a deep model. 82% -> set confidence thresholds. uncertain or novel cases go back to the deeper model. 100% -> verify and measure. check the result, then compare total loop time, cost and errors against your baseline. Jev handles the repeated choices. The deeper model handles hard problems. Verification catches bad routes before they become bad actions. I visualized the full Jev Engineering loop below ↓

Morlex @0xMorlex

a week ago

x.com/i/article/2101…

12 7 60 76K 99