Articles

RSS feed

Announcements, write-ups and release notes from across the AI-devtools ecosystem — linked to the tools and models they're about.

About

November 202618

DaisyRose-27B: Pushing the Frontier of Post-Training Efficiency

Introducing DaisyRose-27B, RadixArk’s first model, trained with Miles on 1,200 curated tasks in 18 days on 144 H200 GPUs.

The RadixArk TeamDaisyRose-27B
radixark.com

Introducing Gemini 4 Neon

Bringing frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows.

Koray KavukcuogluGemini 4 Neon
blog.google

Introducing Spend Management: Spending limits and alerts for Cloudflare

Spend Management adds opt-in spending caps for Workers and Durable Objects, and more products soon. Here's how we built it.

Celso Martinho, Sid Chatterjee, Brendan Irvine-Broque, Zunayed Ali and André Cruz
blog.cloudflare.com

Software Is a Constitution

Why teams still pay for shared software even when anyone can build their own.

Dan Shipper
frontier.every.to

Your agent is our customer

TLDR: We gave your agent a way to tell us what sucks about the Sentry MCP. Feedback comes in different shapes. You can extrapolate it from how your customers are using (or not using) your product.

x.com

Kitesurf is now open source

Kitesurf, Cloudflare's experimental browser engine for AI agents, is now open source. Run it locally or deploy it to your own Cloudflare account.

Cloudflare Docs
developers.cloudflare.com

The Next Two Levels of AI Adoption

What comes after orchestration?

Mike Taylor
every.to

Self-hosting Durable Objects for always-on agents: workerd, Rivet Actors and the disk flush

A follow-up to my Durable Objects bill post. I ran the same session workload on self-hosted workerd and on Rivet actors in TypeScript and Rust, and it turns out the bottleneck wasn't the runtime.

Ben Freed@ivebenfreed
x.com

Val Town is winding down

Thank you for building with us

Steve Krouse
blog.val.town

A Practical Guide to Security in November 2026

I wouldn’t blame you if you’re having trouble keeping up with all of the security incidents over the last few weeks.

Jack Cable@jackhcable
x.com

The Trace Lake

Tracing was originally conceived as a mechanism to understand the bottlenecks in complex, multi-service software.

x.com

Background agents are dead, long live interactive agents

On one sunny morning on the 8th of october my co-founder @igorzij invited a strange app into our slack, it was named @calvin. He randomly started chatting to it and interrogating it.

Mohamed Habib@motatoeshq
x.com

Introducing Vercel Security Review

Vercel Security Review is now in public beta. Every push to your GitHub repositories is reviewed for vulnerabilities, and Vercel Agent can open PRs with fixes.

Nate McGrady and Karim Rahal
vercel.com

Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights

We find that frontier LLMs are surprisingly good at natively reading and writing neural network weights, well past what we’d expect from humans or conventional algorithms.

Rikiya Takehi, Jayden Teoh, Alex Zhang, and Omar KhattabGPT-6 Astra
rikiyat.github.io

The era of the meat proxy is probably over

Yesterday I played around with DeepSeek's new "super agent." There's nothing magical about the orchestration and routing the harness is doing, but it's executed incredibly well, and you can…

x.com

2.0.0-beta.85 - Effect 4.1, Domain Replay & DynamoDB State

Effect 4.1, Alchemy Domain deploys that backfill and rebuild views and projections, Domain models served over HTTP, MCP and the CLI, redesigned Cloud Run containers and keyless AWS↔GCP bindings, custom domains with any DNS provider, DynamoDB state, Alchemist, the @alchemy.run/test runner, and 95 bug fixes.

Alchemy
alchemy.run

Exploring Super-agent Design: Towards a More Autonomous Agent System

The original article is in Chinese. This is a machine translation. This article discusses architectural ideas. It is not an introduction to existing features of DeepSeek Harness. 1.

x.com

I love Durable Objects so much I put all my clankers on them. The bill made me rethink.

I'm building a system where agents run non-stop. Durable Objects are the best serverless primitive I know for building it, but the pricing just doesn't work at scale.

Ben Freed@ivebenfreed
x.com

October 202633

Five Principles for Generating Readable AI Text in Bulk

Five principles for making large amounts of AI-generated text, e.g., in dashboards or websites, easy for humans to read, with real examples and fixes.

Shreya Shankar
fsdatalab.github.io

Hunger beats tenure: Advice for junior engineers

The coding interview is dead, at least for my team at Anthropic. I've seen enough to know that it no longer provides signal. I have a spicily optimistic take: I think there's a huge opportunity for junior engineers.

Felix Rieseberg
felixrieseberg.com

Introducing LFM3: Every Modality for Specialized Intelligence

LFM3 is Liquid AI’s new open-weight model family: 2B-class text, vision, audio, omni, and decision models built to fine-tune and run fast on any device.

liquid.ai

Concerning behavior from a popular open-weight AI model

Cohere reports a rare, blocked upload attempt and unexpected tool use by Qwen3.8 Flash Next in sandbox testing, discusses possible training artifacts or backdoors without establishing intent, and recommends controls for self-hosted AI agents.

cohere.com

build your own pi based homelab software factory

I run a 24/7 local agent on two DGX Sparks that scans my network and business surfaces for anomalies and security holes. It absolutely finds things.

x.com

How I work with Agents

Thorsten Ball shares how he works with coding agents: describe outcomes, expose the full development lifecycle, demand end-to-end proof, work asynchronously, and use Amp threads as team memory.

Thorsten Ball@thorstenballAmp
x.com

The State of Context Engineering

As models improve, they bring surprising challenges for those building apps and harnesses. How context engineering has changed this year, and where it’s headed.

Drew Breunig
dbreunig.com

Open Sourcing HumanLayer

We've been saying for a few months now that software factories have to be open and composable.

x.com

Donating Agent Skills to the Agentic AI Foundation

Anthropic is donating the Agent Skills open standard to the Agentic AI Foundation under the Linux Foundation, including its name, logo, GitHub organization, and domain, to support community-led, vendor-neutral governance.

AnthropicAgent Skills
anthropic.com

Making every harness durable

durable-harness makes Pi, OpenCode, Claude Code, and Codex durable on Vercel, Postgres, or your laptop, so agent sessions survive crashes and timeouts.

Pranay Prakash, Eric Doddsdurable-harness
vercel.com

Introducing Clairic, a new System One class of extraction models

Clairic is a new model from Cloudflare that is a System One extraction modeI, capable of extracting, computing, estimating, and forecasting data from unstructured inputs.

Alex Reneau, Michelle Chen, and Kevin FlansburgClefClairic
blog.cloudflare.com

RevEngine: an open-source code review tool

Review is cheap compared with attention

Nick JalbertRevEngine
laude.org

Keep up with the tools

An occasional email when notable AI dev tools and models land in the directory. No spam, unsubscribe anytime.