process
Articles
Ground It or Flag It: Ending Agent Confabulation
The agent failure to fix is not guessing - it is confident, unmarked inference dressed as fact. The cure separates disclosure from verification.
Read Your Own Codebase First
Prior implementations encode constraints vendor docs omit - check your codebase before any setup guide for a library you have integrated before.
Dead Code Removal Is a Security Audit, Not Just Cleanup
Deleting old adapter code forces a semantic comparison you would not otherwise do - and that is where security gaps surface, not where they get introduced.
Retiring a Fork That Stopped Earning Its Keep
A dependency you control "for safety" can quietly become pure cost. We audited a fork from first principles and found the safe move was to delete it.
Killing Skills on Purpose: A 32-to-24 Retirement Cascade
How a full-day audit cut a 32-skill catalog to 24 - and the operational tests that determined what earned retirement vs. what earned a stay.
Shipping a Cultural Directive Your Agents Read at Session Start
Captain's standing order kept getting forgotten. The fix: commit it to a file agents read at session start - and learn what text belongs there.
When a Behavioral Reflex Becomes a Slash Command
A directive repeated enough times got shipped as a skill. When that is the right call, and when memory or CLAUDE.md would have been the wrong tool.
What Happens When Your AI Agent Briefing Lies
The session briefing said 10 alerts. There were 270. Fixing it took four tracks, $741 in compute, and a captain who refused to accept done.
Design Specs as Agent Infrastructure
Agents building UI from text descriptions produce divergent implementations. Design specs loaded at startup solved this.
What Breaks When You Sprint with 10 AI Agents
Wave-based sprint execution across a fleet of machines works until local state drifts from remote. Here is what we learned.
Multi-Model Code Review - Why One AI Isn't Enough
Why sending code through multiple AI models with different strengths produces higher-confidence findings than any single model.
Multi-Agent Team Protocols Without Chaos
How we coordinate dev agents, PM agents, an advisor, and a human captain using namespaced labels, QA grading, and explicit role boundaries.