Writing
Long-form, practical writing from real engineering work — architecture, AI, business systems and the stack behind them.
AG-UI is an open event protocol for streaming an AI agent to a frontend. What it is, how a run looks on the wire, how it differs from MCP, and when to skip it.
npm 12 blocks dependency install scripts by default, but Node 22, 24 and 26 still bundle npm 10 or 11. My test output, ignore-scripts and the tensorlake case.
A VITE_ or NEXT_PUBLIC_ variable ends up in JavaScript every visitor can read. Here is how an AI key gets into a bundle, how to check your own app safely, the server-side fix, and what to do if it leaked.
A bot check decides who gets in, not how much they spend. A source-checked look at denial of wallet, what OpenAI and Anthropic do and do not cap, and a per-user budget design with a reserve and settle sketch.
AGENTS.md, Agent Skills and MCP are not rival layers. They differ by when they load, what they cost in context and whether anything enforces them. A source-checked decision table and a worked example.
Prompt, context and harness engineering are three scopes of one job. A dated, source-checked timeline of who said what, where the sources disagree, and a way to decide which fix an agent mistake needs.
Cursor, Claude Code and Codex read different instruction files. How AGENTS.md gives them one shared rulebook, and the Claude Code gotcha to avoid.
Several AI systems report perfect IMO 2026 scores. Which were marked by the organisers, which by AI agents, and how one setting moved a score by 14 marks.
The 2026-07-28 MCP spec removes sessions and the handshake. What changed, what breaks, what is deprecated, and how to migrate, checked against the changelog.
A viral post sketched YouTube's backend with no sources. I checked each part against primary documents: what's documented and what's only a guess.
Jev from TypeSafe AI answers questions you define with typed choices and probabilities instead of text. How it works, what it costs and where it breaks.
A harness is everything in an AI agent except the model. Here is what a harness engineer actually builds (tools, context, memory, permissions, budgets, evals) and how to do it well.
You do not need a rewrite to add AI to a Laravel or Next.js application. A step-by-step approach: find the right feature, isolate the model behind a service, validate outputs, handle cost and failure, and ship behind a flag.
Tenancy, billing, background work, environments and the boundaries that let a SaaS product grow without a rewrite. A practical architecture for a first version that is built to last, with Next.js, Laravel and PostgreSQL as the reference stack.
When a generic CRM stops fitting, a custom one is often smaller than people expect. The data model, pipeline design, integrations and reporting that make a custom CRM worth building, and the signs that you should not.
Underneath the hype, an AI agent is a loop: a model chooses a tool, the tool runs, the result goes back in. Here is that loop in detail, the engineering that makes it safe, and how to decide whether you need an agent at all.
Retrieval-augmented generation explained from the engineering side: ingestion, chunking, embeddings, hybrid search, reranking, permissions and citations, and the decisions that determine whether a knowledge base actually answers questions correctly.
A practical architecture for AI applications that do real work inside a business: where the model sits, how it reaches your data and tools, and what has to be true before it goes to production.
A short conversation is usually enough to work out whether the idea needs a prototype, a product or a different approach entirely.