Field notes from systems to agents.

// the record

• Devlog #7: I Ran the Gatekeeper Against My Own Site→Four audit skills inspected The Construct. I fixed 8 findings, declined 2 with evidence, and caught the validator trying to break the CSP it was supposed to protect.July 1, 2026
• Devlog #6: The Product Was Never the Flywheel→I bet twice on a self-improving corpus that gets smarter on its own. I instrumented both bets. Both came back unproven. The thing that survived every honest test was the gatekeeper that refuses to call anything done without independent proof.June 22, 2026
• Devlog #1: What Running 8 Parallel AI Agents Taught Me→The origin story: how single-session AI coding broke at scale, and the architecture patterns that became AgentOps.January 12, 2026
• Why ADHD Brains Are Built for AI Coding→The one system I could never make reliable was my own ADHD brain. AI agents turned out to be the scaffolding I'd needed all along.February 5, 2026
• Reliable AI Agents Need a DevOps Feedback Loop→Bound the work, check it independently, record what passed, and preserve reviewed lessons. Reliable agent delivery looks more like DevOps than magic.May 7, 2026

// method

• 12-Factor AgentOps→November 30, 2025
• The REPL Is Dead. Long Live the Factory.→January 17, 2026
• Multi-Agent Orchestration: Lessons from Running AI Agents Across Dozens of Repos→February 12, 2026
• AgentOps Is Context Orchestration→February 14, 2026
• The Knowledge Flywheel: Beads + an LLM Wiki→February 12, 2026
• Convergent Architecture: Three Domains, One Principle→February 15, 2026
• Spec-Driven Development: Why Your AI Agent Needs a Blueprint→February 12, 2026
• The Validation Bottleneck: Why AI Output Quality Is the New CI/CD→February 12, 2026
• Building vibe-check: 1,563 Lines I Deleted→November 29, 2025

// devlogs

• Devlog #5: When the Platform Catches Up→February 6, 2026
• Devlog #4: Why Your AI Coding Specs Keep Failing→January 29, 2026
• Devlog #3: From Vibe Coder to Vibe Engineer→January 21, 2026
• How to Use Your .claude Directory in Claude Code→January 15, 2026

// for everyone

May 23, 2026Fire for the Rest of UsSame gatekeeper discipline that keeps agents trusted on aircraft: give context, set boundaries, check the work, keep what matches. It works on an inbox too.read the essay →

• When It's Lying→June 15, 2026
• The Gutenberg Moment for Code→December 3, 2025

// personal pattern

• Shadow Integration: What Gaming Taught Me About Reading AI Output→February 12, 2026

// also on the shelf

• Getting Started with Vibe Coding: An AgentOps Quickstart→May 20, 2026
• CDLC: Context Development Life Cycle→May 7, 2026
• From WoW Raids to AI Agents→November 30, 2025
• Building This Website with Vibe-Coding→November 29, 2025

Want these as they land?

New essays go out on the feed. The loop behind them is at /method.