12
14 min read
Context Engineering > Prompt Engineering
Context engineering replaces prompt engineering. Learn to design the full context LLMs receive, with production patterns and real metrics.
Blog
Deep-dive posts on distributed systems, AI engineering, and leadership. Real production insights from 15+ years building at scale.
11
16 min read
Your CI/CD Doesn't Work for AI-Written Code
Tests pass, lint is clean, PR looks perfect. But AI agents introduce failure modes your pipeline wasn't built to catch. Here's what to add.
10
17 min read
4 Agents, 1 Spec: Multi-Agent Orchestration That Works
Multi-agent systems fail when agents do different tasks. They work when agents look at the same thing from different angles. Patterns from production.
09
20 min read
SpecForge: Zero Code, Five Microservices in Parallel
Spec-driven AI development framework. Ship microservices without writing code manually. AST extraction + LLM synthesis patterns inside.
08
19 min read
The Builder-Manager: Code to Understand, Not to Ship
The should-EMs-code debate is a false binary. The builder-manager model survives AI flattening and earns team trust. Here's the third path.
07
12 min read
75% of AI Agents Break Code They Maintain
AI agents excel at new code but fail at maintenance. What actually breaks, why, and the engineering guardrails that make them safe. Tested on real repos.