Articles for people who
ship the stack
Original rewrites on React, Node.js, TypeScript and AI — practical notes from the same engineering practice behind our operator software. Article bodies are in English.
Tagged: ai-coding
Seven Open-Source Tools to Audit AI-Generated Code Before Production
Seven GitHub projects, from mutation testing to load tests, that check test quality, leaked secrets, dependencies, accessibility and speed of AI-built apps.
941 wordsRead articleSpec-First Development: Agree on Intent Before AI Writes the Code
When code is cheap, intent is the bottleneck. A practical spec-driven workflow with OpenSpec, scenarios to review before coding, and tips to roll it out on a team.
1821 wordsRead article14 Claude Code Slash Commands That Will Change How You Work
A practical tour of 14 Claude Code commands, from init, memory and compact to review, doctor and inline shell calls, that cut repetition and keep sessions fast.
1178 wordsRead articleContext, Constraints, Output: A Prompt Template for Reliable AI Code
Vague prompts give vague code. Use a three-part template of context, constraints and output format to get production-ready functions on the first try.
1086 wordsRead articleSix Free Add-Ons That Fix Claude Code's Most Annoying Habits
Six free Claude Code add-ons with install commands: Context7, Playwright MCP, GitHub MCP, document skills, Spec Kit and Sentry MCP, and which to try first.
1450 wordsRead article30 Days of Hands-Off AI Coding: Five Architectural Traps in a Real SaaS
A month of letting AI write all application code for a production micro-SaaS exposed drift, bad webhook handling, debug loops, bloat and a junior-developer problem.
1507 wordsRead articleSetting Up Claude Code in VS Code and Using It Beyond Autocomplete
Install Claude Code, run it from the VS Code terminal, write a CLAUDE.md, and adopt workflows for large tasks, debugging, test-first work and refactoring, plus its limits.
1839 wordsRead articleTen Prompt Engineering Mistakes Developers Make, With Fixes
Ten common prompting errors, from vague requests and missing context to blind trust in output, each paired with a concrete before-and-after prompt you can reuse.
1148 wordsRead articleWhy Your Codebase Is a Graph: Tree-sitter Parsing for an AI Coding Assistant
Why AI coding assistants can't read whole repos, how graphs and syntax trees capture structure, and how to parse code with Tree-sitter to build a symbol index.
1386 wordsRead article
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.