<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Coding Agents on Yue Jiang | 姜岳</title><link>https://yuejiang.org/tags/coding-agents/</link><description>Recent content in Coding Agents on Yue Jiang | 姜岳</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sat, 04 Jul 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://yuejiang.org/tags/coding-agents/index.xml" rel="self" type="application/rss+xml"/><item><title>Eight months into vibe coding</title><link>https://yuejiang.org/2026/07/04/eight-months-into-vibe-coding/</link><pubDate>Sat, 04 Jul 2026 00:00:00 +0000</pubDate><guid>https://yuejiang.org/2026/07/04/eight-months-into-vibe-coding/</guid><description>&lt;p&gt;Since last winter, I started using Claude Code, then Codex, for my work — my current setup is a
Claude Max 5x + ChatGPT Pro. They quickly became something I reach for every day. Below are a couple
of notes on customizations I&amp;rsquo;ve picked up along the way. The first two are for VS Code — I know the
CLI is where the bleeding edge lives and is the best supported, but for the analysis work I do I
prefer being able to quickly interact with files and plots (usually on a remote EC2 instance).&lt;/p&gt;</description></item></channel></rss>