#llm

Articles tagged with #llm

I Built a Self-Improving SDLC Pipeline for Claude Code After a 4-Hour Grafana Bug
A Knowledge Extraction loop that turns each post-incident lesson into a global pattern surfaced in every agent's first step. MTTR ↓ 94% on the second occurrence. Files only, no SaaS.
May 12, 202610 min read
What Separates a Great CTO from an Ordinary One — and Why I Got Tired of Doing It By Hand
Two CTOs answer the same question and end up on different planets. The difference is files in a repo.
May 2, 20268 min read2
I Reverse-Engineered Claude Code and Built a Virtual CTO From the Parts
Lifecycle hooks, model stratification, file-based memory, and seven agents that ship features with two approvals. ~1600 lines, MIT, no SaaS.
Apr 26, 20266 min read5