GARDEN: six principles for agent-first codebases
Introducing GARDEN, six language-agnostic design principles for codebases maintained primarily by LLM agents.
Blog
RSA is dead: Why you should migrate to elliptic curve cryptography today
RSA is slower, larger, and weaker than ECC at every comparable security level. Here is why you should stop using RSA, how to migrate to ED25519, and how to enforce it organization-wide.
Shared CI runners without isolation: What could go wrong
Why sharing CI runners across teams with no resource limits, no cleanup, and no security boundaries is an anti-pattern that causes flaky builds, security incidents, and wasted time.
Manual deployment runbooks: The hidden cost of human-powered releases
Why deployment processes that live in wikis, checklists, or someone's head are one of the most common and costly anti-patterns in software operations.
DNS round-robin is not load balancing
Why using multiple DNS A records to distribute traffic is a reliability anti-pattern, and what to use instead.
Why your CI/CD pipeline is slower than it should be
Common CI/CD anti-patterns that waste developer time, and practical fixes to speed up your pipeline.
Log hoarding without observability
Collecting massive amounts of logs without structured analysis, alerting, or dashboards is one of the most expensive anti-patterns in operations — and one of the easiest to fix.
Stop rebuilding artifacts for every environment
Why rebuilding your application from source for each environment is a reliability and security anti-pattern, and how to fix it with immutable artifact promotion.
TLS re-encryption between internal services: A costly anti-pattern
Why terminating and re-initiating TLS at every internal hop wastes CPU, adds latency, and creates operational headaches — and what to do instead.
Hairpin NAT: The hidden tax on your local traffic
How services connecting to each other via public IP waste bandwidth, exhaust NAT tables, and add latency — and what to do about it.
Infrastructure as Code: Getting started without overengineering
A pragmatic approach to adopting Infrastructure as Code — start small, iterate, and avoid common pitfalls.
Docker in production: Lessons from running containers for years
Practical lessons learned from running Docker containers in production — image hygiene, logging, health checks, and more.