Improving the blog, guides, courses, and tools directory. Daily iterations with Claude Code - most pages you see were built or refined by AI agents.
Creating structured learning paths for Vercel AI SDK, Claude Agent SDK, building CLIs with AI, and agent sandboxing patterns.
Shipping small, useful developer tools. Agent Generator, Migrate (for OpenAI API transitions), and more on the way.
Faceless technical walkthroughs on AI coding tools, agent patterns, and developer workflows. Focus on what actually works in production.
Fair, factual comparison guides for developer tools. Claude Code vs Cursor vs Codex, pricing breakdowns, alternatives guides. SEO-first content that helps people make decisions.
Deep research before every video. Primary sources, current docs, release notes, benchmarks. The research quality determines content quality.
Multi-agent workflows, harness layers, sub-agent parallelization, and how to make AI agents reliable enough for production.
Building software where AI is the primary author. Skills systems, CLAUDE.md conventions, and the emerging patterns of agentic coding.
How to turn one piece of content into many. Video to blog to social to newsletter. Maximizing reach without maximizing effort.
Every sponsored video is a product I actually use. The channel grows slower, but the trust compounds.
Evergreen decision-intent content over news cycle commentary. Comparison posts and pricing guides have longer shelf life than hype coverage.
Nearly everything ships through Claude Code, Codex, or another AI agent. Manual coding is reserved for architecture decisions and debugging.
Working remotely. Building from wherever the internet is fast and the coffee is strong.
Check out the projects, tools, and content that came from these focuses.

New tutorials, open-source projects, and deep dives on coding agents - delivered weekly.