<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Codex on Tokenise</title><link>https://tokenise.rosvetic.com/tags/codex/</link><description>Recent content in Codex on Tokenise</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Sun, 04 Oct 2026 07:00:00 +0000</lastBuildDate><atom:link href="https://tokenise.rosvetic.com/tags/codex/index.xml" rel="self" type="application/rss+xml"/><item><title>Run Codex from a script: codex exec, structured output and CI</title><link>https://tokenise.rosvetic.com/posts/codex-exec-in-scripts-and-ci/</link><pubDate>Sun, 04 Oct 2026 07:00:00 +0000</pubDate><guid>https://tokenise.rosvetic.com/posts/codex-exec-in-scripts-and-ci/</guid><description>&lt;p&gt;Codex isn&amp;rsquo;t only a terminal app you sit in front of. &lt;code&gt;codex exec&lt;/code&gt; runs it without the interactive interface, so it can be a step in a script, a scheduled job or a CI pipeline. The &lt;a href="https://learn.chatgpt.com/docs/non-interactive-mode.md" target="_blank" rel="noopener noreferrer"&gt;non-interactive mode docs&lt;/a&gt;&#10; lay out the behavior, and the defaults tell you how OpenAI expects you to use it.&lt;/p&gt;</description></item><item><title>How Codex reads AGENTS.md, and how to layer it so the right rules win</title><link>https://tokenise.rosvetic.com/posts/codex-agents-md-layering/</link><pubDate>Wed, 30 Sep 2026 17:00:00 +0000</pubDate><guid>https://tokenise.rosvetic.com/posts/codex-agents-md-layering/</guid><description>&lt;p&gt;Codex reads instructions from &lt;code&gt;AGENTS.md&lt;/code&gt; files before it starts work, and it doesn&amp;rsquo;t read just one. It builds a chain of them, in a documented order, and the order decides which rule wins when two disagree. If you&amp;rsquo;ve ever had Codex ignore a rule you were sure you&amp;rsquo;d written down, the layering is the first place to look.&lt;/p&gt;</description></item><item><title>Stop telling your AI agent to 'fix the bug'. Make it reproduce it first</title><link>https://tokenise.rosvetic.com/posts/reproduce-first-bug-prompt/</link><pubDate>Wed, 30 Sep 2026 12:00:00 +0000</pubDate><guid>https://tokenise.rosvetic.com/posts/reproduce-first-bug-prompt/</guid><description>&lt;p&gt;&amp;ldquo;Fix the login bug&amp;rdquo; is a terrible prompt for a coding agent, and a frustrating one to debug when it goes wrong. The agent will read some code, find something plausible, change it, and tell you it&amp;rsquo;s fixed. Sometimes it is. Sometimes it has patched a bug you didn&amp;rsquo;t have.&lt;/p&gt;</description></item></channel></rss>