Deep Thoughts on ByF (BaiFan) · Daily Deep Think

Rethinking Team Division of Labor

blog@baifan.site (ByF (BaiFan)) — Mon, 03 Aug 2026 10:00:00 +0800

The conclusion first, so you don’t waste your read

We used to split teams by craft: frontend, backend, product design. Now we split teams by stage: Prototyper — grabs the first idea and fails fast Builder — turns the prototype into a product you can take to market Maintainer — holds the fort once the software reaches scale Scaler — takes a proven product and scales it 10x, 100x Finisher — polishes the product and the code, sands off the burrs As for why we’re only splitting this way now — the answer involves AI. Read on.

The Future of AI Coding Tools: From Copilot to Autonomous Agent

blog@baifan.site (ByF (BaiFan)) — Mon, 06 Jul 2026 09:00:00 +0800

Twelve articles, from the Think-Act-Observe loop all the way down into the 500,000-line engineering abyss. Now, standing at the end of the series and looking back over the whole evolutionary path, a bigger question surfaces:

What will AI coding tools ultimately evolve into?

A better Copilot? A more powerful IDE plugin? Or something new that we haven’t named yet?

Three Evolutionary Paths
#

Looking back at the development of AI coding tools, you can see three clear stages of evolution:

3,000 vs 500K Lines: The Architectural Abyss From Toy to Product

blog@baifan.site (ByF (BaiFan)) — Thu, 02 Jul 2026 09:00:00 +0800

The claude-code-from-scratch project implements a “mini Claude Code” in roughly 3,000 lines of TypeScript. It has the Think-Act-Observe loop, a tool system, a system prompt, permission control, memory, Skills, multi-agent, MCP — every core concept we tore down over the previous ten posts.

And the official Claude Code source? Roughly 500K lines of TypeScript/TSX.

That’s a 100x+ gap between 3,000 lines and 500K. What’s in it?

The gap isn’t “number of features”. from-scratch already covers every core concept. The gap is the road each concept travels from “it runs” to “it’s usable” — that road is called productionization.

The Discrete State Machine of Plan Mode: Decision Tree Pruning, Backtracking Costs, and Diminishing Returns

blog@baifan.site (ByF (BaiFan)) — Tue, 16 Jun 2026 09:00:00 +0800

Autoregressive large language models operating in unconstrained environments suffer from an inherent tendency toward myopic mutation and runaway divergence. Claude Code’s Plan Mode attaches an external discrete Finite State Machine (FSM) that projects the action space onto a strictly side-effect-free subset, using an interactive human barrier to prune catastrophic physical backtracking branches from the decision tree.

In disciplined software engineering, experienced developers instinctively value structural planning: mapping dependency graphs, deriving type signatures, and scoping blast radiuses before modifying a single line of code. Large language models possess no such innate restraint. Shaped by reinforcement learning to generate immediate answers, an agent granted unrestrained write permissions will instinctively mutate files without global context, precipitating an architectural cascade across complex repositories.

The MCP Protocol: USB-C for AI Tool Interconnection

blog@baifan.site (ByF (BaiFan)) — Fri, 12 Jun 2026 09:00:00 +0800

Before MCP, every AI tool integration was a custom build. Want ChatGPT to reach a database? Write a plugin. Want Claude to call the GitHub API? Write a Function Calling definition. Want Copilot to access your internal systems? Write a Custom Tool.

Every time, the same thing: define an interface, write adapter code, handle auth, manage errors. The wheel, reinvented countless times.

The problem MCP (Model Context Protocol) solves is simple: give AI agents a standard protocol for connecting tools. Just as USB-C gave electronic devices a universal connector, MCP gives AI agents a universal tool interface.

Multi-Agent Architecture: The Art of Fork-Join Collaboration

blog@baifan.site (ByF (BaiFan)) — Tue, 09 Jun 2026 09:00:00 +0800

There’s a counterintuitive phenomenon: take the same model, split it into multiple independent instances, and have them collaborate — the result is often better than a single instance.

Not more compute (three instances burn more total tokens than one), not more capability (every instance is the same model). The gain comes from role separation — when each agent attends to only one facet of the problem, it works with more focus and more depth than a “do-everything agent”.

The Nature of Skills: Reusable Encapsulated Expertise

blog@baifan.site (ByF (BaiFan)) — Fri, 05 Jun 2026 09:00:00 +0800

If you’ve used Claude Code, you may have noticed the /skills command listing installed “skills”. Writing standards, architecture design methods, TDD workflows, code review strategies — it sounds like a plugin marketplace.

Skills are not plugins.

That’s the first and most important thing to understand about the Skill system. A plugin is code — it extends what the program can do. A Skill is a prompt — it extends what the model knows.

AI Memory Systems: Four Memory Types and Semantic Recall

blog@baifan.site (ByF (BaiFan)) — Tue, 02 Jun 2026 09:00:00 +0800

Last time we covered context compaction — solving “how much the current session can remember”. But a coding assistant needs another kind of memory: long-term memory across sessions.

You close Claude Code and reopen it tomorrow. What should it know?

You prefer TypeScript over JavaScript
You said last time “don’t use the underscore library”
This project uses Vitest, not Jest
You once corrected one of its misunderstandings

Context compaction can’t handle these. The session ends, the history clears, tomorrow is a fresh window. If the AI needs you to re-explain everything each time, it isn’t an assistant — it’s a burden.

The Physical Cost of Context Compression: KV Cache Pressure, Attention Decay, and Lossy Summarization

blog@baifan.site (ByF (BaiFan)) — Sat, 23 May 2026 09:00:00 +0800

Large language models possess no biological memory at the hardware tier; they operate entirely on Key-Value Caches that scale linearly in GPU High-Bandwidth Memory (HBM) alongside attention weight distributions that mathematically dilute as sequence length grows. The notion of long-horizon conversations presented by Claude Code is, at its engineering core, an eviction and compaction pipeline operating within the tight margins of memory capacity and numerical precision, trading away information entropy and prefix caching efficiency.

Permissions and Real Security Boundaries: Sandboxes, System Calls, and Approval Escape

blog@baifan.site (ByF (BaiFan)) — Wed, 20 May 2026 09:00:00 +0800

Any AI agent permission system running in userspace that relies on model self-discipline and application-level regex filtering does not constitute an enforceable security boundary from the perspective of the operating system kernel. Claude Code’s 5-level permission model and interactive approval workflow operate as flow-control protocols to mitigate developer mistakes, lacking sandbox containment against adversarial attacks.

Granting an AI coding assistant filesystem write access and terminal execution rights under the belief that confirmation prompts preserve a safety perimeter is a fundamental category error. Confusing interactive human approval with privileged access control collapses into a cosmetic barrier when confronted with supply-chain poisonings, indirect prompt injections, and shell parser differentials.

System Prompt Engineering: Decisions Behind 800 Lines

blog@baifan.site (ByF (BaiFan)) — Sat, 16 May 2026 09:00:00 +0800

If you think a system prompt is just “giving the AI a persona”, Claude Code’s system prompt will upend that notion.

It isn’t a few sentences. It’s a thousand-line template engine made of multiple layers, each responsible for something different: identity definition, behavioral constraints, tool documentation, project context, memory injection, skill loading. Some layers are static (identical every time); others are dynamic (injected in real time based on the project, the session, user preferences).

Tools Are Capability: AI Agent Tool System Design Philosophy

blog@baifan.site (ByF (BaiFan)) — Tue, 12 May 2026 09:00:00 +0800

Last time we talked about the Think-Act-Observe loop — the skeleton of an AI coding assistant. A skeleton can’t work on its own; it needs muscle. In Claude Code, that muscle is tools.

Without tools, the loop loses the Act between Think and Observe and degenerates into plain Q&A. With tools, the model can touch the real world: read your code, change your configs, run your tests.

The Soul of an AI Coding Assistant: The Think-Act-Observe Loop

blog@baifan.site (ByF (BaiFan)) — Thu, 07 May 2026 09:00:00 +0800

If you dig through the Claude Code source, you will discover a counterintuitive fact:

The core of the entire program is a while loop.

Not a complex state machine, not an elaborate pipeline, not distributed coordination — just a plain while loop that does three things over and over: let the model think, let the model act, feed the results back.

1
2
3
4
5
6
7
8
9


while (user hasn't quit) {
 thinking = callLLM(conversation history + tool definitions)
 if (thinking.wants to call a tool) {
 result = executeTool(thinking.toolCall)
 conversationHistory.push(result)
 } else {
 output(thinking.reply)
 }
}

That’s it? The entire secret of an AI coding assistant lives in this loop?

AI Psychosis: When MTTR Fanaticism Meets Resilience

blog@baifan.site (ByF (BaiFan)) — Tue, 24 Feb 2026 14:00:00 +0800

I have a strong intuition: some companies in the tech industry right now are going through a full-blown AI psychosis, and inside those companies, rational conversation is simply impossible. I’m not talking about strangers — I’m talking about personal friends I deeply respect. That leaves me worried about how this ends.

This fear has a specific source. I lived through the great MTBF (mean time between failures) vs MTTR (mean time to recovery) debate when infrastructure moved from traditional ops to cloud and cloud automation. Now those exact same arguments are resurfacing in a new costume — only the scope has widened from infrastructure to the entire software industry, and possibly to the whole world.

A New Playbook for AI Programming: UI First, Then the PRD

blog@baifan.site (ByF (BaiFan)) — Tue, 17 Feb 2026 14:00:00 +0800

Everyone knows the standard Vibe Coding flow by now: write user stories → produce a PRD → AI generates the backend → AI generates the frontend → demo and verify. It works, but at delivery time the UI interaction experience often falls far short of expectations. Where does it go wrong? In the PRD itself.

1. The Fundamental Problem with PRDs
#

PRDs (Product Requirements Documents) were never that friendly to humans in the first place. Three reasons:

Immortality Is Just Eternal Loneliness

blog@baifan.site (ByF (BaiFan)) — Fri, 05 Dec 2025 20:00:00 +0800

Immortality Is Just Eternal Loneliness
#

Today I saw a piece in the NT saying that China’s longevity industry is advancing at full speed and that people living to 150 is just around the corner. I searched the Chinese domestic news, and every story said people would only live to 120 — which turned out to be the press sparing the honored any embarrassment. Awkward.

Not Even Your Life Is Wanted: The Lost Bargaining Chip in the Machine Age

blog@baifan.site (ByF (BaiFan)) — Tue, 04 Nov 2025 20:00:00 +0800

The presentation slides at tech keynotes are always immaculate. Bathed in cool hues of cyan and clean white, tailored executives gesture smoothly beneath the spotlights, unveiling high-frame-rate b-roll of six-axis robotic arms dancing across cleanrooms and unmanned shearers cleaving coal seams in the bowels of the earth. The official press releases are draped in immaculate corporate dignity: “Zero-Harm Production,” “Automated Hazard Mitigation,” “Upgrading the Workforce Through Intelligent Infrastructure.”

Applause ripples through the auditorium. Commentators celebrate a triumph of humanitarian progress: human beings will no longer be crushed under collapsing rock faces, nor will they gamble their marrow against the freezing crosswinds of high-rise scaffolding just to secure their next meal.

AI Uncertainty and Carpe Diem: An Ordinary Person's Thoughts

blog@baifan.site (ByF (BaiFan)) — Tue, 28 Oct 2025 20:00:00 +0800

After reading the recent magazines and reports, my honest advice is this: while these next couple of years are still livable, seize the day and go enjoy yourself.

Two assumptions:

Suppose AI really does keep breaking through and takes over most people’s jobs. Society will inevitably enter a turbulent period of painful adjustment, and nobody knows what will happen.
Suppose AI fails to break through and, in the end, cannot create that much value. Given the enormous leverage the whole world has piled onto AI today, the fallout when the bubble bursts would be no smaller than the 2008 subprime mortgage crisis. Nobody knows what will happen.