How I gradually removed confirmation gates, added behavioral and architectural guardrails, and learned to let Codex CLI practise TDD autonomously.
When I Entrusted TDD to an AI Agent
How I gradually removed confirmation gates, added behavioral and architectural guardrails, and learned to let Codex CLI practise TDD autonomously.
I built an "agent" last year that embarrassed me. Not because it failed. Because it succeeded, and...
An edit to 64 columns in two layers ended most thinking loops: 62% right vs 8% on one RTX 4090. What to measure and which settings matter.
Most LLM wiki setups I've seen are very good at one kind of memory and have almost none of the...