Sho Naka

Sho Naka

Jul 17

I Almost Built a VSCode Extension. Four GitHub Searches Changed My Mind.

#cli #opensource #productivity #ai

1

9 min read

Sho Naka

Jul 15

I Built 3 Checks Into My AI's Slack Workflow. I've Used One of Them Zero Times.

#ai #mcp #slack #automation

7 min read

Sho Naka

Jul 14

The Humans Become the Bottleneck: A Structural View of AI-Augmented Teams

#aiagents #teamdesign #humanintheloop #operations

5 min read

Sho Naka

Jul 14

CLAUDE.md Is Not a Prompt File. It Is an Operating Boundary.

#ai #claude #contextengineering #devops

1

5 min read

Sho Naka

Jul 12

I Measured Claude Code's Prompt-Cache Cost Three Ways. 85% of It Wasn't Mine to Trim.

#claudecode #anthropic #llm #vscode

4 min read

Sho Naka

Jul 11

McDonald's ran an AI drive-thru pilot for 3 years across 100+ stores, then killed it. It wasn't an accuracy problem.

#aiadoption #digitaltransformation #productivity #genai

4 min read

Sho Naka

Jul 10

I Stopped Asking AI Agents to Behave: 3 Permission Levels Instead

#ai #aiagents #devops #automation

2

7 min read

Sho Naka

Jul 9

Open-ended delegation freezes half a room. Here's what unfroze them

#leadership #mentoring #management #productivity

9 min read

Sho Naka

Jul 7

My menu bar clock disappeared after a macOS update. I wrote a 50-line VS Code extension instead of trusting a marketplace one.

#vscode #security #macos #javascript

5 min read

Sho Naka

Jul 7

Guard Scripts Don't Stop an AI Agent From Publishing Your Draft. It Has a Shell

#ai #claudecode #security #automation

8 min read

Sho Naka

Jul 7

Stop Fixing Your AI Writing Prompt. Make These 5 Decisions First

#ai #writing #productivity #promptengineering

11 min read

Sho Naka

Jul 6

Don't Trust Your SSH Config Until ssh -G Proves It: The Host * Ordering Trap

#ssh #devops #security #beginners

6 min read

Sho Naka

Jul 5

LLM-as-Judge Is Too Lenient. Here's a Cheap Fix: Judge Refute (Maybe) Arbitrate

#ai #llm #architecture #testing

1

8 min read

Sho Naka

Jul 4

Why Skipping Research and Structure Costs More When You Know the Topic Well

#writing #productivity #softwaredesign #career

4 min read

Sho Naka

Jul 4

My Mac dev environment broke on every brew upgrade. Here's the shared-dependency bug I kept fixing wrong.

#homebrew #python #macos #devops

3 min read

Sho Naka

Jul 1

Why is the review getting stuck even though it became faster with the introduction of AI?

#ai

10 min read

Sho Naka

Jun 30

Choosing Between Image AI and Code-First Tools for Technical Diagrams: An 8-Tool Comparison

#illustrations #mermaid #python #plantuml

1

5 min read

Sho Naka

Jun 27

Tests Pass, Design Breaks: Why TDD Can't Hold the Line on Design Intent

#tdd #testing #softwaredesign #aidevelopment

1

7 min read

Sho Naka

Jun 26

My AI agent built a feature that already existed 15 times

#ai #agents #productivity #devops

4 min read

Sho Naka

Jun 26

AI made generation cheap. It did not make judgment cheap.

#ai #productivity #leadership #devops

1

5 min read

Sho Naka

Jun 24

No errors, but the email never arrives — it's probably not your code

#php #beginners #email #webdev

5 min read

Sho Naka

Jun 23

I Copied a Google AI Studio Session by Hand. Most of the Data Was Gone.

#aitools #googleaistudio #gemini #dataloss

1

3 min read

Sho Naka

Jun 23

Why AI Keeps Making the Same Mistake — And Why Correcting It Each Time Doesn't Work

#aicollaboration #claude #promptengineering #rlhf

1

6 min read

Sho Naka

Jun 22

Why 'Just Be Careful Next Time' Never Reaches an AI

#aicollaboration #aiagents #promptengineering #humanintheloop

1

6 min read

DEV Community

Badges

Writing Debut

I Almost Built a VSCode Extension. Four GitHub Searches Changed My Mind.

I Built 3 Checks Into My AI's Slack Workflow. I've Used One of Them Zero Times.

The Humans Become the Bottleneck: A Structural View of AI-Augmented Teams

CLAUDE.md Is Not a Prompt File. It Is an Operating Boundary.

I Measured Claude Code's Prompt-Cache Cost Three Ways. 85% of It Wasn't Mine to Trim.

McDonald's ran an AI drive-thru pilot for 3 years across 100+ stores, then killed it. It wasn't an accuracy problem.

I Stopped Asking AI Agents to Behave: 3 Permission Levels Instead

Open-ended delegation freezes half a room. Here's what unfroze them

My menu bar clock disappeared after a macOS update. I wrote a 50-line VS Code extension instead of trusting a marketplace one.

Guard Scripts Don't Stop an AI Agent From Publishing Your Draft. It Has a Shell

Stop Fixing Your AI Writing Prompt. Make These 5 Decisions First

Don't Trust Your SSH Config Until ssh -G Proves It: The Host * Ordering Trap

LLM-as-Judge Is Too Lenient. Here's a Cheap Fix: Judge Refute (Maybe) Arbitrate

Why Skipping Research and Structure Costs More When You Know the Topic Well

My Mac dev environment broke on every brew upgrade. Here's the shared-dependency bug I kept fixing wrong.

Why is the review getting stuck even though it became faster with the introduction of AI?

Choosing Between Image AI and Code-First Tools for Technical Diagrams: An 8-Tool Comparison

Tests Pass, Design Breaks: Why TDD Can't Hold the Line on Design Intent

My AI agent built a feature that already existed 15 times

AI made generation cheap. It did not make judgment cheap.

No errors, but the email never arrives — it's probably not your code

I Copied a Google AI Studio Session by Hand. Most of the Data Was Gone.

Why AI Keeps Making the Same Mistake — And Why Correcting It Each Time Doesn't Work

Why 'Just Be Careful Next Time' Never Reaches an AI