Blog
Writing about AI-assisted development, productivity workflows, and lessons from building with AI agents every day.
Two Tools for Seeing What Your Agent Built
Automated screenshots for visual verification and a desktop capture tool — the two ways I check what my AI agents produce.
Custom Status Lines in Claude Code: Knowing Which Instance You're In
When running 4 Claude Code instances across 2 machines, a custom terminal status line with context tracking, identity, and cost keeps you oriented.
Refactoring a Monolith: Extracting a 3PL Module Into Its Own App
How we extracted 27 API endpoints, 14 database tables, and 7 handler modules from one app into another — with zero downtime and a sprint-based approach.
How Plan Mode Changed the Way I Build Software
Plan Mode in Claude Code forces alignment before code is written. Here's how this single workflow pattern has transformed my development process and why I consider it indispensable.
What It Actually Looks Like to Build Software with AI Agents
4 Claude Code installations, 2 machines, up to 6 simultaneous sessions. A real look at daily AI-assisted development — the actual setup, workflow, and what makes it produce quality code.
Why I Built Claw Recall: AI Agents Need Memory
Every AI conversation is temporary. Context compaction erases what your agent was just working on. Claw Recall fixes this — a searchable memory system for AI agents.