https://blog.margrop.net/en/tag/openai/ Recent content in Openai on Margrop Blog Hugo -- gohugo.io en-US Sun, 24 May 2026 08:10:00 +0800 https://blog.margrop.net/en/post/codex-reset-radar-quota-reset-guide/ Sun, 24 May 2026 08:10:00 +0800 https://blog.margrop.net/en/post/codex-reset-radar-quota-reset-guide/ <blockquote> <p><strong>The short version</strong></p> <p>Codex resets have recently become more important to watch. Sometimes the reset is part of the normal usage cycle. Sometimes it is a compensation reset after an incident. Sometimes it follows a fix for abnormal limit consumption. For heavy Codex users, the painful part is not that the quota resets. The painful part is discovering that you still had useful weekly quota left, but you did not use it before the reset overwrote the old window.</p> <p><code>Codex Reset Radar</code> is a small public dashboard that turns scattered public signals into a practical warning: if a Codex quota reset window is likely or already open, it tells you to use your remaining weekly quota on valuable work before it expires.</p> </blockquote> <p>This article discusses only public pages, public status signals, and public references. It does not include private accounts, tokens, internal systems, private network details, or personal usage data. Any discussion of quota, reset windows, and early warnings is a user-side workflow analysis, not an official statement about OpenAI billing, subscription policy, or service guarantees.</p> https://blog.margrop.net/en/post/codex-130-approval-flags/ Tue, 12 May 2026 18:50:00 +0800 https://blog.margrop.net/en/post/codex-130-approval-flags/ <blockquote> <p><strong>The short version</strong></p> <p>If you upgraded to Codex 0.130.0 and noticed that the old <code>/approvals</code> habit no longer works, the practical fix is not to downgrade. Move the permission policy to launch time. Use <code>--ask-for-approval</code> to control when Codex asks, <code>--sandbox</code> to control what commands can touch, and reserve <code>--dangerously-bypass-approvals-and-sandbox</code> for environments that are already isolated outside Codex. The long flag can also be written directly as <code>--yolo</code>, which is easier to remember, easier to type, and more consistent with command habits from tools such as Gemini.</p> <p>In other words, stop treating approval mode as an in-session toggle. Treat it as part of the way you start the agent.</p> </blockquote> <p>This article only discusses public Codex CLI flags and generic workflows. It does not include real hostnames, internal addresses, usernames, tokens, private paths, or project names. Commands use placeholders such as <code>&lt;PROJECT&gt;</code> and <code>&lt;TASK&gt;</code>.</p> https://blog.margrop.net/en/post/openai-gpt-5-4-introduction-2026/ Fri, 06 Mar 2026 08:20:00 +0800 https://blog.margrop.net/en/post/openai-gpt-5-4-introduction-2026/ <p>On March 5, 2026, OpenAI officially released GPT-5.4. The first question many people asked was simple: is this just another small point release, or is it a real shift in how much productive work the model can finish?</p> <p>If you only look at the version number, 5.4 looks like the next iteration after 5.2 and 5.3. But once you line up the official launch notes, model docs, API capability matrix, and the ChatGPT-side product updates, the core value of GPT-5.4 is not &ldquo;how many more parameters it has.&rdquo; The real change is that it pulls three lines of capability into a much more usable center point:</p> <ol> <li>Reasoning</li> <li>Coding and engineering execution</li> <li>Tool ecosystem coordination, including computer use</li> </ol>