博客

I've been shipping what people are now calling 'AI-enabled web development' for about a year - local-first coding agents on top of Ollama, Claude Code plugins, MCP servers wired into real codebases, ONNX models running in the browser. Here's what's worked, what hasn't, and the loops I keep coming back to.

Building a tokenizer from the ground up - UTF-8 byte streams, character-level encoding, and Byte-Pair Encoding explained through working code.

Training a spam classifier, then systematically breaking it with perturbed inputs - a hands-on walkthrough of adversarial examples in NLP.

Exploring event-driven architecture, the event loop, and why non-blocking I/O outperforms thread-per-request models.

Step-by-step deployment with Node 22, PM2, Nginx, HTTPS via Certbot, and automated deployment aliases.

A security incident report - detection, forensics, and hardening after finding xmrig on a compromised VPS.

Quick reference for finding your IP address across macOS and Linux using ifconfig, ip, and related networking commands.

Phrase-based password generation combining familiarity with entropy for credentials that are both memorable and secure.

Password security fundamentals - brute-force attacks, dictionary attacks, and supplementary measures like 2FA.

Remote debugging iOS Safari pages using Mac's Web Inspector - console, elements, network, and breakpoints.

V8 engine internals - parsing, Ignition bytecode, TurboFan compilation, and deoptimization triggers.

GPT-4o's multimodal capabilities - native training across text, voice, vision, and real-time latency improvements.

Text-to-image mechanics - diffusion models, text encoding, iterative denoising, and current limitations.

Leveraging the HTML autocomplete attribute to build smarter, faster forms with zero extra JavaScript.

Express.js meets OpenAI API - building a personality-driven chatbot with system prompt engineering.

Running classic arcade games inside a Docker container - a fun intro to containerization and port mapping.

Core EC2 concepts - instance types, AMIs, security groups, and getting started with AWS compute.

A practical guide to writing Dockerfiles - FROM, RUN, COPY, multi-stage builds, and best practices.

Techniques for making JavaScript-heavy apps search-engine friendly - SSR, lazy loading, and structured data.

Achieving a perfect Lighthouse score in Nuxt - image optimization, code splitting, and performance tuning.