Discover

Curated articles from sources across the web

37410 articles from 16 sources
The Context Packet: The Right Architectural Coding Agents
D
DEV Community · Technology

The Context Packet: The Right Architectural Coding Agents

A coding agent with access to your entire repository does not necessarily understand your system better. It may understand less. Because a large repository may contains obsolete code, and decisions that were never written down. Giving an agent more files increases the amount of text it can inspec...

Majdi Zlitni
Read
It Passed Every Test. That's Why It Can't Ship Yet.
D
DEV Community · Technology

It Passed Every Test. That's Why It Can't Ship Yet.

My order-reading LLM passed the 29-question exam. Zero fatal errors. The model is chosen. So — ship it? No. And the reason is the most important thing in this series. The exam has holes. I am the holes. The questions, the answer key, the product catalog — I made all of it. Which means...

John Green
Read
test
D
DEV Community · Technology

test

testing updating canonical url without deleting tags.

Nick Taylor
Read
Our AI reviewer invented a request. Our producer retried 245 times.
D
DEV Community · Technology

Our AI reviewer invented a request. Our producer retried 245 times.

We run ~100 LLM agents unattended on local models. Last week we found one document that had been rewritten 245 times in 5 days — every attempt rejected. A sibling document: 225 times. Combined, about 470 wasted generations, all burned on the same two files. Here is the autopsy, with the actual numb...

GX Cafe LLC
Read
The Model That Costs 3x More Won by Exactly One Question
D
DEV Community · Technology

The Model That Costs 3x More Won by Exactly One Question

I gave the same 29-question order-reading exam to two models. A cheap one (Haiku 4.5) and one that costs about three times as much (Sonnet 5). The result Cheap model 28 of 29 questions clean Expensive model all 28 executed questions clean Fatal errors zero for both...

John Green
Read
Predicting Shareholder Meeting Failure: How AI Quantified Risk from a Single Sentence in Past Minutes and a 5x Increase in Share
D
DEV Community · Technology

Predicting Shareholder Meeting Failure: How AI Quantified Risk from a Single Sentence in Past Minutes and a 5x Increase in Share

Hey, it's your friendly neighborhood dev-grandpa here, still chugging along building AI agents on weeknights and weekends. Today, I want to share a story about how one of my custom analysis bots dug up a significant risk from a completely unexpected angle. I'll walk through how the AI quantitativel...

oji - building AI in public
Read
Construyendo un recomendador de emparejamiento de expertos
D
DEV Community · Technology

Construyendo un recomendador de emparejamiento de expertos

La forma del problema Un directorio es una superficie: el miembro lo abre y adivina. Un recomendador es una superficie de empujar: el sistema propone y tiene que justificarse. La justificación es la parte difícil, y es donde vive la estadística. Tres restricciones hicieron esto distinto...

Franchesco Romero
Read
ESP32 + Python: From Microcontroller to IoT
D
DEV Community · Technology

ESP32 + Python: From Microcontroller to IoT

ESP32 + Python: From Microcontroller to IoT Artcal 0: Introduction When it comes to transferring data from one place to another, things can sometimes become tricky, especially when communication happens between the hardware and software levels. In this article series, I would love to share the exp...

Janada Pelahara
Read
Log Level Design and Rotation: Why Apps Use DEBUG/INFO/WARNING/ERROR
D
DEV Community · Technology

Log Level Design and Rotation: Why Apps Use DEBUG/INFO/WARNING/ERROR

Open any application log and you'll see the same kind of message tagged with different labels: DEBUG, INFO, WARNING, ERROR. Why not just write down everything that happens, in one uniform stream? This post looks at what log levels actually do, and at the companion problem every long-running app even...

Susumu Takahashi
Read
Same tokens. Different DNA. Why a Texas notebook must not become a parlor.
D
DEV Community · Technology

Same tokens. Different DNA. Why a Texas notebook must not become a parlor.

Most agent "redesigns" are mood boards. Navy glass. A fake dashboard. The dates and proper nouns already in the file never make it onto the page. /reimagine-it is the opposite contract: extract this source's nouns, dates, and colors, then ship a real artifact (webpage, infographic poster, SVG, Thre...

K Merchant
Read
How we run five coding agents side by side in one window
D
DEV Community · Technology

How we run five coding agents side by side in one window

I build NestMux, a desktop app that runs Claude Code, Codex, Gemini CLI, Copilot and OpenCode in a grid of real terminals, each with its own account and its own git worktree. This is how it works inside, including the parts that are heuristics wearing a confident UI and the parts that do not survive...

Eliseo Fernandez Suarez
Read
I pushed the badBANANA Threat Observatory public
D
DEV Community · Technology

I pushed the badBANANA Threat Observatory public

This was one of those projects I had wanted to make for a long time, even before I had a very clear reason for it. Over the years, if you spend enough time around security, you keep seeing threat maps. A lot of them are visually impressive. They pull you in immediately. Lines moving across a globe,...

GnomeMan4201
Read

Reading List