Công cụ b4 dành cho các nhà phát triển kernel Linux đang thử nghiệm giao diện người dùng dạng văn bản (TUI) để hỗ trợ kiểm tra mã bằng trợ lý AI như Claude Code. Mới đây, công cụ này đã tự review chính các bản vá của mình — một bước tiến thú vị. Tính năng là tùy chọn, giúp tăng hiệu suất và phát hiện lỗi ẩn, nhưng không bắt buộc sử dụng. #b4 #LinuxKernel #AIReview #ClaudeCode #OpenSource #Linux #KernelDevelopment #AIAssistant #DevTools #CôngCụPhátTriển #NhânLinux #TrợLýAI

https://www.phoronix.c

Btw: I made a thing (tbh: I'm still in the middle of doing it)... 😃

#Claude #ClaudeCode #Terminal #Swift #TerminalUI #TUI

https://oldbytes.space/@LAYERED/115991509116265447

LAYERED (@LAYERED@oldbytes.space)

Attached: 1 image Sneak preview: I'm in the process of making something. If you like the look and feel of terminals, if you like terminal-based tools/apps, if you're a software developer and simply want to have terminal UIs in your software... well, I'm building something really great. 😃 I've hidden an Easter egg on the landing page. Once you've found it, wait a moment... there's a second one. 😎 #Terminal #TUI #Swift #TUIkit #macOS #Linux https://tuikit.layered.work

OldBytes Space - Mastodon

Boris Cherny (@bcherny)

/config에서 'Explanatory' 또는 'Learning' 출력 스타일을 활성화해 Claude Code가 변경의 '이유'를 설명하게 하고, Claude에게 익숙하지 않은 내용을 설명하도록 시각적 HTML 프레젠테이션을 생성하게 하는 등 Claude Code를 학습 용도로 활용하는 몇 가지 팀 팁을 안내합니다.

https://x.com/bcherny/status/2017742759218794768

#claude #claudecode #learning #ai

Boris Cherny (@bcherny) on X

10. Learning with Claude A few tips from the team to use Claude Code for learning: a. Enable the "Explanatory" or "Learning" output style in /config to have Claude explain the *why* behind its changes b. Have Claude generate a visual HTML presentation explaining unfamiliar

X (formerly Twitter)

[TypeScript 10만 줄을 Rust로, Claude Code 실전 포팅기

전 Facebook 엔지니어 Christopher Chedeau가 Claude Code를 활용해 10만 줄의 TypeScript 코드를 Rust로 포팅한 실험 프로젝트. 목표는 Pokemon 배틀 AI 학습용 고속 오라클 구축으로, 1개월 만에 3.5배 속도 향상과 99.96% 일치율 달성. 차분 테스트와 명확한 지시가 성공의 핵심 전략이었으며, AI 코딩 도구의 현주소와 한계를 평가한 사례.

https://news.hada.io/topic?id=26295

#typescript #rust #claudecode #codeporting #aiassistant

TypeScript 10만 줄을 Rust로, Claude Code 실전 포팅기

<p>전 Facebook 엔지니어 Christopher Chedeau (Vjeux)가 <strong>Pokemon Showdown</strong> 배틀 엔진(약 10만 줄 TypeScript)을 <strong>Claude...

GeekNews
Got access to #claude and #claudecode at work this week. Holy crap, I didn't realize how good AI had gotten. It's not perfect, but it's successfully helped me build some tools for work and a few simple tasks I've been meaning to automate. Being able to feed it the context of internal tools and have it learn and write fresh code is really cool. #ai

Jarred Sumner (@jarredsumner)

지난 24시간 동안 PR로 Claude Code의 콜드 스타트 시간이 40% 개선되고 메모리 사용량이 32%~68% 감소하는 성능 최적화가 이루어졌다는 업데이트입니다. 아직 목표 수준은 아니지만 개선이 진행 중임을 알립니다.

https://x.com/jarredsumner/status/2017521250965590058

#claude #claudecode #performance #optimization

Jarred Sumner (@jarredsumner) on X

In the last 24 hrs, the team has landed PRs to Claude Code improving cold start time by 40% and reducing memory usage by 32% - 68%. It’s not yet where it needs to be, but it’s getting better.

X (formerly Twitter)

Der Testdrang von Claude fasziniert mich immer wieder.

#claude #claudecode

NASA just trusted an AI with planning part of a real Mars rover drive

Perseverance completed two successful AI-planned drives on Mars.

https://gadgetbond.com/nasa-perseverance-claude-ai-planned-mars-rover-drive/

Anthropic expands Cowork with open-source agent plug-ins

With new plug-ins, Anthropic’s Cowork app enables reusable AI workflows for sales, legal, finance, data, and other knowledge-heavy roles.

https://gadgetbond.com/anthropic-cowork-agentic-plugins/

Anthropic’s Claude Code now shows how much AI really writes your code

AI pair programming now comes with numbers engineering leaders can trust.

https://gadgetbond.com/anthropic-claude-code-contribution-metrics-analytics-public-beta/