AI Tools on Mi&Bee Blog

The Hidden Trap of Headless Browsers: Why Can't Your Automation Tool Catch Early Page Errors?

Fri, 15 May 2026 00:00:00 +0000

Introduction

You’re debugging a frontend engineering issue — the page is behaving abnormally. You ask an AI to open the page with a browser tool and check the console for errors.

The AI opens the page, scans around, and tells you: The console is clean, no errors whatsoever.

You’re skeptical. You open Chrome DevTools yourself — three bright red errors are staring you in the face, the page has already crashed into a white screen. The AI visited the exact same page using a Headless browser, so why did it catch nothing?

One Month with the Zhi Theme: Mermaid v11 Upgrade Experience

Wed, 13 May 2026 00:00:00 +0000

One Month In

It’s been nearly a month since the previous article Switched My Blog Theme: From Hugo NexT to Self-Written Zhi. During this month, the theme has been running stably without major issues. Replacing the NexT theme was the right decision — although there was some initial adjustment, the experience is now significantly better.

After a month of use, the theme’s stability has exceeded expectations. My initial concerns — whether pure Hugo Pipes without build tools could support complex requirements — were proven unfounded. Daily maintenance has become very simple; modifying a feature no longer requires digging through deeply nested SCSS files — one CSS file gets the job done.

Evolution: Oh My OpenAgent Configuration Iteration Log

Thu, 07 May 2026 00:00:00 +0000

The previous article covered the initial configuration setup. This one documents the adjustments after two weeks of running: expanding from single vendor to a four-tier model pool, adding fallback chains, hitting the GLM-4.5-air trap of analyzing without writing code.

This post covers: fallback strategy design, complete free model pool inventory and analysis, concurrency control configuration, and the decision process for GLM-4.5-air replacement.

After the previous article’s initial configuration, I ran it for two weeks — all the issues that needed fixing surfaced.

Switched My Blog Theme: From Hugo NexT to Self-Written Zhi

Thu, 16 Apr 2026 00:00:00 +0000

Why Switch

This blog previously used Hugo NexT, forked for custom modifications. NexT itself is a feature-rich theme, but when it comes to “customizing things yourself,” the experience wasn’t great.

The issues boiled down to a few things:

SCSS nesting hell. 101 SCSS files, three levels of directory nesting. _common/components/post, _common/components/third-party, _common/outline/sidebar… To change a style, you first had to figure out which file it was in, where variables were defined, and which scheme was overriding it. Not that it couldn’t be done, but each change meant half an hour of hunting.

Zhipu Coding Plan × Oh My OpenCode: Multi-Model Orchestration Setup Guide

Sun, 05 Apr 2026 00:00:00 +0000

Why Bother

When it comes to writing code with AI, the gap between single-model and multi-model approaches keeps widening. No matter how strong a single model is, it can’t compete with a team of specialized models working in parallel.

Oh My OpenCode (OmO for short) is a multi-model orchestration plugin in the OpenCode ecosystem, with 11 Agents each having distinct responsibilities and 48 Hooks spanning the entire lifecycle. Zhipu’s Coding Plan provides access to the full GLM model series. Combining the two allows you to assign different models by role — strong coders for coding, strong reasoners for reasoning, free models for busywork.

Domestic LLM Resource and Cost Comparison: GLM-5 / Kimi K2.5 / MiniMax M2.7

Mon, 23 Mar 2026 00:00:00 +0000

Overview

This article compares the resource requirements and usage costs of three major domestic LLMs, helping developers choose the right solution for their scenarios.

Model	Vendor	Architecture	Minimum Deployable VRAM	API Available
GLM-5	Zhipu AI	Dense (multiple versions)	24GB (8B)	✅
Kimi K2.5	Moonshot AI	MoE (undisclosed)	24GB (lightweight)	✅
MiniMax M2.7	MiniMax	MoE 230B	Not yet open-sourced	✅

GLM-5 (Zhipu AI)

Versions & Hardware Requirements

GLM-5 offers 4 parameter versions, making it the widest-coverage domestic LLM currently available.