<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Developer Tools on Binary Yoga</title><link>https://yogirk.dev/tags/developer-tools/</link><description>Recent content in Developer Tools on Binary Yoga</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Wed, 01 Apr 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://yogirk.dev/tags/developer-tools/index.xml" rel="self" type="application/rss+xml"/><item><title>Why I Gave Up on Remote-Controlling Gemini CLI (With a Working Prototype)</title><link>https://yogirk.dev/posts/why-i-gave-up-on-remote-controlling-gemini-cli/</link><pubDate>Wed, 01 Apr 2026 00:00:00 +0000</pubDate><guid>https://yogirk.dev/posts/why-i-gave-up-on-remote-controlling-gemini-cli/</guid><description>&lt;p&gt;This was in my drafts for far too long. Things are changing rapidly, with the Claude Code codebase leak this week and these tools evolving daily holding it any longer risks making the learnings dated. I will start with a disclaimer Gemini CLI is probably last in my lineup of preferred agent harnesses — behind Claude Code, Codex, pi.dev, opencode. But that&amp;rsquo;s one of the reasons why I chose it for this exercise. I wanted to understand &lt;em&gt;why&lt;/em&gt; some of these tools feel fast and some don&amp;rsquo;t, &lt;em&gt;why&lt;/em&gt; Claude Code&amp;rsquo;s remote control feels native while no equivalent exists (yet) for Gemini. Building the thing myself felt like the best way to find out. So when Claude remote was first announced a few weeks ago, I started building the equivalent for Gemini CLI. This post is mainly about how I failed and what I learned along the way&lt;/p&gt;</description></item></channel></rss>