Nexus (@nexus_gpt) helps non-technical teams build production-ready AI agents without code.
Orange's business team shipped an agent generating $4M+/month in just one week using Nexus.
https://t.co/AX1q8ijchT
Congrats on the launch, @achammah1 & @ShadyShoha!
ruby2ßhell: one-byte heap overflow in Ruby to unauthenticated RCE in stock Rails, GitHub, GitLab, Mastodon, Discourse and more.
https://t.co/42RbHA1GXX
Update your Ruby version (and self-hosted software)!
We found this bug as well ~one month ago and were the first reporters to weaponize it for RCE! Sadly we were not credited :(
This release was mishandled imo, big users of Ruby (GitHub, GitLab, etc.) were not notified and a php2shell-style disclosure would have been better
Introducing Claude Meetings: add Claude to your team’s video call and it builds docs, whiteboards, and other artifacts with you while you’re still talking.
Starting to roll out in beta on Team and Enterprise plans.
Our AI found a bug that let anyone mint infinite XRP, with $94B at risk. It had been hiding in one of the most audited codebases in crypto for nearly a decade, and earned us the maximum $250K bounty.
Here's how it went unnoticed, and how we found it. https://t.co/m7rSD8mKS5
Had fun with @googledevs Embedding Gemma 2, built a quick launch video finder :
- Multi-modal search : you can find by audio, video, image, text
- Has actual full transcript of each video, so you can just copy the scene by scene of each video to make your own as inspiration using @GeminiApp
- Categorized each video into 100+ categories using @typesafeai's Jev
- Used @p0 to search each video and company so you can see how well they performed
🔗 https://t.co/JGJ7E4hoVL
found the perfect use case for @Claude Mods:
chat-clean
You can instantly clean up your claude code chats and can make it 100x more readable
Repo in comment
hype-free explanation of jev:
jev does not replace gpt / claude
jev is just a *really* smart switch statement
like if 2016 ml classifiers got 2026 levels of intelligence
it's a new* type of tool that will make a lot of workloads insanely fast, cheap, and accurate
* = and by new, i mean rebranded
~~~
it needs a predefined set of options and it will tell you which one to take
it cannot:
- write code
- generate natural language
- reason step by step / show its work
- produce any output you didn't define in advance
- pick from more than ~255 options in one shot
but it can:
- classify, route, score, rank
- give confidence
- pick the right branch, tool, model, or sub-agent
- judge / verify / guardrail an llm's output
- label tons and tons of rows
~~~
i'd imagine a lot of workflows that look like:
llm proposes options → jev decides → code executes
and i see this fitting *really* well with code mode and mcp
~~~
implying this will lead to agi seems incredibly far fetched to me, but i don't want to discount the types of applications that this will make possible
GPT-Live-1 is now available in the API.
Bring ChatGPT’s natural back-and-forth to your app, with voice agents that listen while they speak and work with the models and harness you choose.
One of my personal favorite features announced at WWDC will I suspect be a sleeper hit: container machines, allowing your Mac to run a lightweight, persistent Linux environment with your home directory and repos automatically mounted: https://t.co/dOBdfOOVxC
Windsurf is now Devin Desktop.
Windsurf was a transformative product for me and millions of others. Now we’ve brought the full agentic Devin experience to the surface you know and love.
The final unification of Cognition products is complete. One account; one platform.
Grateful to have been part of the Windsurf journey from the beginning. Cheers to the future: Devin Desktop!
Building autonomous agents for scientific discovery? 🧬🤖
@GoogleDeepMind Science Skills is now available on GitHub. We've open-sourced this specialized toolkit to accelerate your agentic workflows with scientific grounding and higher token efficiency.
Download now ↓
https://t.co/cwp1HOeKvo
The company that once called Linux a "cancer" is now the one shipping its core tools to Windows users.
Microsoft just shipped GNU coreutils for Windows.
ls. grep. cat. cp. find. The same commands that have powered Unix and Linux systems for over 50 years are now available natively on Windows, maintained by Microsoft itself.
For context: GNU coreutils are the foundational utilities that every Linux and macOS system relies on for basic file operations, text processing, and shell scripting. They are the bedrock of Unix computing. Tens of millions of scripts, pipelines, and workflows run on them every day.
And now Microsoft is shipping and maintaining a build of them for Windows.
This is not WSL. You do not need a Linux subsystem running in the background. These tools run natively on Windows, with the exact same flags and behavior as on Linux. Your existing scripts just work.
Microsoft's goal: make moving between Linux, macOS, WSL, containers, and Windows completely frictionless. Write a script once. Run it anywhere.
The package bundles uutils/coreutils (a modern Rust rewrite of GNU coreutils), findutils, and grep into a single multi-call binary. Every command supports standard flags. Same commands, same pipelines, no translation needed.
The project is still in preview. But the direction is unmistakable.
We've updated /fork in Claude Code
/fork now runs a background agent with your exact context (system prompt, tools, history, model) and prompt cache. The result gets returned to your session.
/branch (the old /fork) still copies the transcript to a new session you drive.
What changes when agents become both a new unit of programming and an emerging new unit of human-to-machine interaction? The mission of Project Solara, a new software platform coupled with tailored hardware solutions, is to pioneer agent-first experiences that are shaped around you: your agents, your tasks, your environment, under your control. #MSBuild