knes on Hacker News

1

A good AGENTS.md is a model upgrade. A bad one is worse than no docs at all (opens in new tab)

(augmentcode.com)

6knes18d ago0

2

Designing with code, without writing code (opens in new tab)

(augmentcode.com)

2knes1mo ago0

3

How I went AI-native in my terminal workflow (opens in new tab)

(augmentcode.com)

2knes1mo ago0

4

Review the Spec. Not the Code (opens in new tab)

(augmentcode.com)

5knes1mo ago0

5

Building faster than you can think (opens in new tab)

(augmentcode.com)

1knes2mo ago0

6

Agent's context is a junk drawer (opens in new tab)

(augmentcode.com)

2knes2mo ago0

7

Why every AI coding breakthrough feels normal within 90 days (opens in new tab)

(augmentcode.com)

1knes2mo ago0

8

Why every AI coding breakthrough feels normal within 90 days (opens in new tab)

(augmentcode.com)

2knes2mo ago0

9

Spec isn't cool. You know what's cool? 10 specs (opens in new tab)

(augmentcode.com)

1knes2mo ago0

10

Show HN: We got sick of juggling terminals for AI agents so we built a workspace (opens in new tab)

(augmentcode.com)

1knes2mo ago0

11

Show HN: We got sick of juggling terminals for AI agents so we built a workspace (opens in new tab)

(augmentcode.com)

3knes2mo ago0

12

Rebuilding state management: How we made our VS Code extension 2× faster (opens in new tab)

(augmentcode.com)

1knes4mo ago0

13

Why GPT-5.2 is our model of choice for Augment Code Review (opens in new tab)

(augmentcode.com)

12knes5mo ago4

14

We Added ACP Support to Augment (Now in Zed, Neovim, and Emacs) (opens in new tab)

(augmentcode.com)

3knes6mo ago0

15

Developers are choosing older AI models (opens in new tab)

(augmentcode.com)

183knes6mo ago176

knes

Recent submissions

A good AGENTS.md is a model upgrade. A bad one is worse than no docs at all (opens in new tab)

Designing with code, without writing code (opens in new tab)

How I went AI-native in my terminal workflow (opens in new tab)

Review the Spec. Not the Code (opens in new tab)

Building faster than you can think (opens in new tab)

Agent's context is a junk drawer (opens in new tab)

Why every AI coding breakthrough feels normal within 90 days (opens in new tab)

Why every AI coding breakthrough feels normal within 90 days (opens in new tab)

Spec isn't cool. You know what's cool? 10 specs (opens in new tab)

Show HN: We got sick of juggling terminals for AI agents so we built a workspace (opens in new tab)

Show HN: We got sick of juggling terminals for AI agents so we built a workspace (opens in new tab)

Rebuilding state management: How we made our VS Code extension 2× faster (opens in new tab)

Why GPT-5.2 is our model of choice for Augment Code Review (opens in new tab)

We Added ACP Support to Augment (Now in Zed, Neovim, and Emacs) (opens in new tab)

Developers are choosing older AI models (opens in new tab)

Recent submissions

A good AGENTS.md is a model upgrade. A bad one is worse than no docs at all (opens in new tab)

Designing with code, without writing code (opens in new tab)

How I went AI-native in my terminal workflow (opens in new tab)

Review the Spec. Not the Code (opens in new tab)

Building faster than you can think (opens in new tab)

Agent's context is a junk drawer (opens in new tab)

Why every AI coding breakthrough feels normal within 90 days (opens in new tab)

Why every AI coding breakthrough feels normal within 90 days (opens in new tab)

Spec isn't cool. You know what's cool? 10 specs (opens in new tab)

Show HN: We got sick of juggling terminals for AI agents so we built a workspace (opens in new tab)

Show HN: We got sick of juggling terminals for AI agents so we built a workspace (opens in new tab)

Rebuilding state management: How we made our VS Code extension 2× faster (opens in new tab)

Why GPT-5.2 is our model of choice for Augment Code Review (opens in new tab)

We Added ACP Support to Augment (Now in Zed, Neovim, and Emacs) (opens in new tab)

Developers are choosing older AI models (opens in new tab)