13 articles exploring claude code
Skills, subagents, and MCP servers aren't tiers of the same capability - they solve three different context-cost problems. A primary-sourced decision framework.
74% of AI coding tool users report meaningful productivity gains. Most designers are not in that number. The blocker is not skills or fear. It is project structure. A complete designer-ready Claude Code repo with DESIGN.md, design tokens, skills, agents, and a weekend setup guide.
Most tools marketed as knowledge graphs for AI coding are dependency graphs. I built three production KGs, ran the experiments on LangChain, and the results were not what the vendors claimed.
Production AI systems are 98% harness, 2% model. New research reveals why architecture, permissions, and safety matter more than model capability - and how to build systems that actually work.
AI writes 80% of my code. I still review 100% of these 5 file types. A blast-radius framework ranking what to review line-by-line, and what to trust.
Smaller, constrained AI models force clarity and structure. I build faster with Haiku than Opus because constraints eliminate bad habits. Here's why.
Most AI coding tool comparisons still reward the wrong things. A workflow-first breakdown of Claude Code, Cursor, Copilot, Windsurf, and Antigravity through the lens that actually matters: how teams ship under real constraints.
AI-generated code feels fast, but the maintenance cost appears later. Why AI creates locally correct but globally fragile systems, and the engineering standards that fix it.
Why one-off prompting does not compound, and how to move from isolated prompts to repeatable AI workflows using playbooks, MCP data sources, and action layers.
A battle-tested blueprint for Claude Code projects that stay predictable at 50+ files - folder layout, .claudeignore, skills architecture, and 3 ready-to-copy templates.
The MCP servers that matter most for real AI leverage: analytics, email, calendar, GitHub, databases, observability, SEO, social, docs, and file storage. Plus practical playbooks for turning them into repeatable workflows.
The 10 Claude Code skills that now separate developers who merely generate from those who ship differentiated products. From UI taste and frontend structure to brand systems and skill creation.
Battle-tested techniques for .claudeignore, context compression, and multi-agent task splitting that cut token usage by 60–90% without losing Sonnet's predictive accuracy.