The Rookie Weekly Vol.001: Rebooting My System by the Nanming River, 10 Pitfalls of Digital Life

💡 Late-Night Review: The Window to Fight Entropy

The wind from the Nanming River in Guiyang sneaks through the window cracks, carrying a damp, herbal scent. I stare at the time in the bottom-right corner of my screen, 03:14. The 27-inch 4K monitor in front of me emits a faint blue glow, reflecting off the rim of my now-cold Luckin Coffee cup. A ring of dried brown stain marks the bottom, much like my fragmented sleep over the past two weeks.

As a full-stack developer working from Huaguoyuan, I’m constantly battling “entropy.” Server logs overflow inexplicably, and my codebase grows into something resembling over-fermented dough as requirements pile up. At the start of 2026, I stared at the algorithmically curated junk feed on Twitter and felt a wave of nausea.

If I don’t build my own “filter,” someone like me will eventually drown in the flood of information. The sole reason for XBSTACK (an AI productivity system) is that I personally built this fortress in the digital wasteland.

🧩 Underlying Architecture: Reflections on Abandoning Next.js

When refactoring the entire site, I committed the greed every programmer falls for: wanting to use the most impressive, cutting-edge tech. I initially built a version using Next.js 15, only to realize it was sheer “over-engineering” for a content-focused site like mine.

Why I Chose Astro 5?

Next.js made me feel like I was maintaining a heavy tank, while Astro gave me a sharp scalpel.

Next.js forces the browser to load 200KB of React runtime by default, even when I just want to display plain text. It’s a resource hog. Astro, on the other hand, operates on an “0KB by default” philosophy, injecting JS only where interactivity is needed (like that calculator).

While using Astro 5’s Content Layer, I hit that dreaded ERR_CONTENT_LAYER_MISSING error. I sat in front of my computer for three hours troubleshooting, only to discover it was a permissions issue with the cache directory. I manually ran rm -rf .astro, watching the console resume its compilation progress, and slowly regained that sense of control.

Hardcore for Beginners

🚀 AI Workflow: Stop “Slicing,” Try MCP + Cursor

This week, I was refactoring a quantitative backtesting script. The sheer shock of the Model Context Protocol (MCP) still makes my hands tremble when I think about it.

Previously, when using AI to write code, I had to frantically copy-paste documentation to feed it. Now, I wrote a local MCP Server. My 14-inch MacBook Pro’s fans spun up, emitting a faint whistling sound. Through the protocol, Cursor could directly “see” my local 1.2GB SQLite database schema.

I wrote a snippet of connection code:

import mcp
from mcp.server import Server

app = Server("xiaobai-db-tools")

@app.call_tool("query_sqlite")
def query_sqlite(sql: str):
    import sqlite3
    conn = sqlite3.connect("my_quant_data.db")
    return conn.execute(sql).fetchall()

AI is finally no longer “blindly guessing” my logic. It’s “auditing” my engineering. My error rate on complex SQL queries dropped from 30% to 0. This is the physical dominance of “context awareness.”

📉 Leek Review: “Cash Defense” Amidst Nasdaq New Highs

The Nasdaq 100 (QQQ) hit a new all-time high again. The so-called “gurus” in my WeChat Moments are banging gongs and drums once more. I (a novice) look at the floating profits in my account, yet I feel a chill down my spine.

Why I Still Doggedly Hold SGOV?

In my 80/20 barbell strategy, 20% of my position is permanently locked in SGOV (0-3-month US Treasury ETF). Its current price is 100.45 USD.

Many people mock me, asking if a 5% return even counts as investing? All I can say is: in a crazy bull market, cash isn’t trash; cash is an expensive “put option.”

SGOV is T+0 level ammunition. On the day the market pulls back by 10%, those who went all-in will be wailing, while I’ll use SGOV dividends to pick up the cheapest chips.

This week’s live trading: Not a single share of tech stock added. Full position bought into SGOV for risk avoidance. Only by protecting your principal do you earn the right to talk about compounding.

🛠️ Toolbox: My Craftsman’s Box This Week

1. Exa.ai: I’ve completely turned off Google Search. Exa returns only high-quality papers and hardcore blogs.
2. HHKB Professional Hybrid: Typing on it feels like cutting through a block of premium butter. My charcoal-gray HHKB has developed a sheen on the spacebar, a medal of honor from 200 million keystrokes. 👉 (Don’t buy cheap ones; your fingers deserve the best.)

🧠 Wealth Logic

The prerequisite for compounding is ensuring you survive until the inflection point arrives.

The essence of this newsletter is to solidify my one-time thinking costs into digital assets that can be distributed infinitely. Next week, I’ll walk you through how I used code to generate my first bit of “passive income.”

📬 Next Week’s Preview

Vol.002: The First Pot of Gold — How I Used Python to Automate Arbitrage Opportunities.

---

• Wealth Algorithm: Optimize your risk-reward ratio via 2026 Asset Allocation Draft.

• Wealth Algorithm: Optimize your risk-reward ratio via 2026 Asset Allocation Draft.

• Practical Tools: Use my developed Compounding Calculation Engine for stress testing.

💬 Comment Interaction

1. What is the most headache-inducing bug you’re currently facing in development?
2. What potential performance bottlenecks do you see in this technical solution?
3. If it were you, which alternative approach would you choose to implement this?
4. Has this layout worked out in your real life? Did you encounter any permission pitfalls?