Posts
12
Articles collected into this reading track.
Posts
12
Articles collected into this reading track.
Cadence
Opinionated field notes on how strong software teams stay clear, calm, and accountable.
Publishing rhythm for this series.
Focus areas
4
Recurring themes shaping the articles in this track.
This series turns a private set of notes into public writing. The common thread is simple: teams get faster when they stop treating quality, communication, and delivery as separate conversations.
Some of these posts are about code. Some are about reviews, releases, and the habits around the code. Together they describe the standards I keep coming back to when I want a team to scale without becoming noisy, fragile, or dependent on heroics.
Focus areas
Read these in order if you want the clearest progression from foundations to deeper implementation decisions.
Part 1
Why strong engineers care about product outcomes, users, and system health instead of stopping at ticket completion.
In series: The santi020k way · Part 1
Part 2
The code smells that usually predict rising maintenance cost before a team feels the full pain.
In series: The santi020k way · Part 2
Part 3
Replace ad-hoc string literals with clearer domain language, stronger typing, and safer refactors.
In series: The santi020k way · Part 3
Part 4
A lightweight review language that makes feedback clearer, kinder, and easier to act on.
In series: The santi020k way · Part 4
Part 5
Simple commit and pull-request habits that reduce review friction and make collaboration calmer.
In series: The santi020k way · Part 5
Part 6
Responsive work gets easier when breakpoints and layout decisions are treated like a shared system.
In series: The santi020k way · Part 6
Part 7
The version of coding standards I would actually hand to a growing frontend team today.
In series: The santi020k way · Part 7
Part 8
A release process should lower stress, tighten feedback loops, and make production changes easier to trust.
In series: The santi020k way · Part 8
Part 9
Positive conditionals usually reduce mental load and make branches easier to scan, review, and change.
In series: The santi020k way · Part 9
Part 10
AI speeds up code changes, but ESLint, tests, snapshots, and end-to-end checks are what keep probabilistic output from turning into production risk.
In series: The santi020k way · Part 10
Part 11
AI agents rerun checks constantly. Quiet logs, compact reporters, and parallel verification reduce token waste without lowering the quality bar.
In series: The santi020k way · Part 11
Part 12
DX improvements that cannot be measured rarely survive long enough to compound. Treating them like product work changes that.
In series: The santi020k way · Part 12