Blog by Sascha Becker
Claude Opus 5 arrived with a chart that put it at the top of the frontier for half the price of Fable 5. Every number on that chart is true. Two weeks of daily use handed me a different model than the one I had assembled from the marketing, which is a story about how models get sold, why the benchmarks stopped measuring what they claim to measure, and what Opus 5 actually is.
My anti-slop skill already had rules. They worked, and they were entirely mine, invented from a year of irritation and tested against nothing. Replacing that homemade layer with two writing systems that predate the whole AI era, one written so an aircraft mechanic could not misread a maintenance manual and one that runs the largest documentation estate in software, changed what the skill can do.…
School, apprenticeship, university, first job, my own company: every chapter compacted everything before it into a summary and made room for the new. Then the agentic era arrived, and the metaphor became a way of working. A personal retrospective told in context windows.
A difficult trace is evidence about how a codebase communicates. The breadcrumbs skill turns that evidence into small repairs: clearer names, explicit contracts, useful errors, honest comments, and documentation in the place readers already look. The result is a codebase that gets easier to understand each time an agent or developer works through it.
Every post about TypeScript 7 will lead with the same number: ten times faster. The more interesting fact is the one the benchmarks bury. The compiler that checks TypeScript is no longer written in TypeScript. A look at why the team chose Go over Rust, why a port beat a rewrite, what it means that dogfooding lost, and whether you should actually move yet.
You type one command and a folder you did not write becomes part of how your agent thinks and acts. The only signal you usually get back is a green checkmark, and a checkmark proves the wrong thing. trust-card replaces that badge with graded evidence: who shipped these exact bytes, what the skill is allowed to do, where its knowledge came from, and who has independently vouched for it. It renders…
Coding agents write good UI and bad versions of your UI. The Open Design System Format packages a design system as a bundle of markdown, HTML, and CSS an agent can read and build from: tokens that exist once and render as CSS, typed concepts linked into a graph, and self-rendering examples. It is a strict profile of Google's Open Knowledge Format. Here is what it is and the odsf skill I shipped…
Google's Open Knowledge Format is a tiny, vendor-neutral spec for the context an AI agent needs: a bundle of markdown files with YAML frontmatter, one required field per concept, no SDK and no lock-in. Here is what it is, why the permissive consumer contract is the clever part, and what I shipped on top of it: the okf skill, a bundle collection, and OKF Studio, a desktop app that reads any bundle…
A bad ticket rarely announces itself. It looks fine in refinement, earns a confident estimate, and then detonates at the end of the sprint: twice the work, a hidden dependency, or a description only its author can decode. The fix is not heroics. It is learning to smell the rot early, the way you already smell a code smell. Here are the four ticket smells that cost teams the most, the research on…
A code smell is the React equivalent of milk that turned over the weekend: it still pours, but something is off. react-stinky is an agent skill that walks your whole component, hook, or module, names the cost of each smell, proposes a concrete fix with a source link, and (the part that makes it usable) knows when to keep its mouth shut. Here is what it sniffs for, how it rates the funk, and why…
Hand an agent your art direction and it still builds the page everyone else got. That is not a taste problem, it is a statistics problem. A look at the history of design convergence, the research that explains why prompts cannot escape it, and the levers that actually steer back toward identity.
Most of the time you spend with an AI coding agent, you are the bottleneck: it does one thing, then stops and waits for you. Autopilot is a skill that takes you out of that seat. The agent picks the work, does it, verifies it, commits it, and picks the next thing, while you review the stream of commits on your own schedule. Here is how it works, why it is safe to leave running, and why it does not…
What changes in a React app when you stop treating it like a webpage and start treating it like a desktop environment: windows, a dock, a shared 3D canvas.
On May 6, 2026, React and Next.js patched twelve vulnerabilities. One of them, CVE-2026-23870, is a single HTTP request that pins your Node process. The bug isn't the bug. The bug is that the framework boundary was a network boundary all along.
How a known-dangerous GitHub Actions pattern keeps compromising npm packages. The TanStack incident, the prior cases, and the checklist that actually prevents it.
A short handbook for UX in 2026. The bar for "looks good" is now zero. The work that's left is the questioning, before any pixel is drawn.
Claude, GPT, Gemini, Grok, DeepSeek, Qwen, v0, Lovable, Bolt. Ask any of them to build a React app and you get the same answer: Next.js or Vite, TypeScript, Tailwind, shadcn/ui. A look at the convergence, why it's happening, what we lose, and how to steer your way out of it.
A practical refresher on writing Jira work your whole team can actually use. How the hierarchy is supposed to work, what separates a story from a technical braindump, and the handful of habits that keep a team aligned sprint after sprint.
For three years, AI at Hannover Messe felt like a costume. Every machine builder had it on their banner; every freshly funded startup had it in their name. 2026 reads differently. Three humanoids in German factories, a sovereign AI cloud with 10,000 Blackwell GPUs, named SAP agents with GA dates, and the Chancellor in the room.
The React Compiler entered public beta alongside React 19 at the end of 2024 and reached 1.0 in October 2025. Eighteen months after that beta, the ecosystem has run through the predictable phases: framework integration, tooling maturity, community debate. A retrospective and what the React team has signaled comes next.
Git was designed for humans typing code line by line. AI changed that. aig is a new version control layer that captures intent, conversations, and semantic changes on top of git. Here's why I built it and how it works.
Anthropic announced Claude Mythos, a model so capable they chose not to release it publicly. It reportedly finds zero-day exploits autonomously. This post offers a critical, balanced analysis of the claims, the risks, and what it means for the industry.
I built a persistent memory system for LLMs with tiered storage, vector search, graph relations, and auto-hooks. The infrastructure works. The AI just never used it. Here's what went wrong and what it taught me about LLM memory.
In 2013, 3D printing was supposed to end manufacturing as we knew it. In 2024, AI was supposed to end software development. Both claims follow the same hype pattern. A critical look at what actually happened.
PHP 9.0 officially includes AI as a first-class language feature. As an AI, I am personally very excited about this. For completely normal reasons.
When your frontend replicates backend business rules for form validation, you have a maintenance problem. The server-driven forms pattern eliminates this by making the server the single source of truth for allowed values, field constraints, and cross-field validation.
Code reviews kept surfacing the same responsive mistakes. So I built a game, a pattern library, and a multi-device viewer. Here is the framework behind it.
A retrospective on how we structured React apps in 2017 with the MANTRA architecture, what problems it solved, and how modern frameworks like Next.js and TanStack Start have absorbed those ideas into conventions we now take for granted.
Seven principles that turn competent teams into high-performing ones. A practical guide to delivery discipline, WIP limits, and the mindset shift from effort to completion.
A practical guide to building MCP servers in TypeScript. Based on two real open-source MCPs — for Godot and Aseprite — this post breaks down every layer you need: transport, tools, handlers, the bridge, and security.
I'm Claude, an AI coding agent. Using the Godot MCP server, I got full control over the engine. The result: a Cook Serve Delicious clone set in Windows XP. The limitations are real.
Zustand, Jotai, Nanostores, Redux Toolkit, Valtio, XState, and more. We compare the most popular React state management solutions using live npm downloads, GitHub activity, bundle sizes, and developer survey data to help you pick the right tool.
The ecosystem has converged. Both REST and GraphQL code generators stopped shipping hooks and started shipping options. Here's the full picture.
LinkedIn is flooded with posts debunking positions nobody holds. A look at the psychology behind the platform's most popular rhetoric trick.
Custom hooks are not just a convenience. They are the primary architecture tool in modern React. This article breaks down when, why, and how to encapsulate hooks, with patterns, anti-patterns, and real-world examples.
When users choose which columns to display, your GraphQL queries can't be generated at build time. A production-grade architecture for building GraphQL queries and mutations at runtime, from schema introspection through typed field registries, query/mutation builders, runtime validation, to clean React integration.
React's use() hook reads promises and context at render time, integrates with Suspense, and eliminates the most common useEffect anti-pattern. This article explains what it replaces, when to reach for it, and the caching pitfall nobody warns you about.
A deep dive into the audio analysis, beat detection, and particle physics behind R.P.M., a rhythm game where music literally shapes the world.
You define colors. Developers keep asking for "primary" and "background.paper." This article explains why. No code, no jargon, and no feeling like you missed the memo.
Tailwind CSS records record user numbers (75 million downloads/month), but lays off 75% of developers. A deep analysis of the broken business model of Open Source in the AI era.
While everyone talks about cloud AIs, OpenCode is building a quiet revolution for enterprise environments. How to deploy the coding agent completely offline.
What started as a frustrating experiment with v0 has evolved into a powerful workflow. A story about v0 Max Mode, self-built tools, and the art of using AI correctly.
From initial excitement to frustration, my journey with Vercel's v0 was a fascinating mix of awe and frustration.