What Is Adler Analysis in Stage 0 of the Cangjie‑Skill Pipeline? A Complete Guide

Adler analysis is the foundational Stage 0 "Whole‑Book Understanding" step in the cangjie‑skill pipeline that requires reading an entire source work and producing a structured four‑part interpretation before any skill extraction begins.

This rigorous methodology—named after the Adler reading‑comprehension method—ensures that every downstream skill is anchored in a thoroughly vetted understanding of the source material. The kangarooking/cangjie-skill repository implements this as a reproducible, template‑driven process that outputs a canonical BOOK_OVERVIEW.md file.

The Four Steps of Adler Analysis

The analysis in methodology/01-stage0-adler.md decomposes into four sequential sub‑steps. Each step builds on the previous, creating a complete cognitive map of the source work.

Step 1: Structural Analysis

This step captures the book's skeleton. The analyst answers:

• What type of work is it? (methodology, biography, philosophy, etc.)
• What is the one‑sentence summary of its core thesis?
• What are the 3‑7 primary arguments and how do they relate? (parallel, progressive, contrast, refutation)
• What central problem is the author trying to solve?

The output feeds the 骨架 (skeleton) and 论点关系 (argument relationship) fields in the template.

Step 2: Interpretive Analysis

Here the analyst translates the author's language into a personal map:

• Identify recurring key terms and record the author's own definition for each
• Re‑state 5‑15 core propositions in your own words
• Outline the argument chain that links propositions to evidence

This corresponds to the 关键术语 (key terms) and 核心命题 (core propositions) sections of BOOK_OVERVIEW.md.template.

Step 3: Critical Analysis

The Adler "critical" phase exposes the author's blind spots before you agree with them:

• Era limitations: When was the work written and which premises may be outdated?
• Perspective blind spots: How do the author's background, industry, or culture bias the argument?
• Unproven assumptions: Which claims are taken for granted without justification?
• Strongest counter‑argument: What would a skeptic most persuasively raise?

This Critical section feeds directly into the Boundary (B) field of each downstream skill, as defined in SKILL.md.

Step 4: Applicability Analysis

A cangjie‑skill extension to the classic Adler method. This step evaluates what can become reusable skills:

• Which sections suit skillification? (frameworks, checklists, principles, decision‑processes)
• Which parts are unsuitable? (pure history, anecdotes, emotional narrative)
• Rough estimate of downstream skill count and priority ranking based on "most empowering for ordinary people"

The BOOK_OVERVIEW.md Output

The final deliverable of Stage 0 Adler analysis is a populated BOOK_OVERVIEW.md file, rendered from templates/BOOK_OVERVIEW.md.template. This document becomes the global context for every subsequent extractor and skill‑generation stage.

The template uses Jinja2‑style variable substitution. Key variables include:

• 结构_类型 — work type classification
• 一句话主旨 — one‑sentence thesis
• 骨架 / 论点关系 — structural elements
• 关键术语 / 核心命题 / 论证链 — interpretive elements
• 时代局限 / 立场盲点 / 未被证明的假设 / 最强反对意见 — critical elements
• 可_skill化内容 / 不适合_skill化内容 / 预估_skill数量 / 优先级排序 — applicability elements

Practical Implementation

Below is a runnable Python example showing how to programmatically generate the Stage 0 overview using the repository's template:

import datetime
from pathlib import Path
import jinja2

# Load the Jinja2 template from cangjie-skill repository

template_path = Path(
    "/__modal/volumes/vo-cSqLfqnnIwYXEonuEJnnZa/repos/github.com/kangarooking"
    "/cangjie-skill/main/templates/BOOK_OVERVIEW.md.template"
)
template = jinja2.Environment(
    loader=jinja2.FileSystemLoader(template_path.parent)
).get_template(template_path.name)

# Stage 0 Adler analysis data gathered from source material

context = {
    "BOOK_TITLE": "Deep Work",
    "AUTHOR": "Cal Newport",
    "YEAR": "2016",
    "DATE": datetime.date.today().isoformat(),
    "SOURCE_FILE": "books/deep-work.pdf",
    # Step 1: Structural

    "结构_类型": "实操手册",
    "一句话主旨": "在高度注意力时代，深度工作是提升价值的关键。",
    "骨架": ["解释深度工作的概念", "展示深度工作的价值", "提供培养深度工作的实践技巧"],
    "论点关系": "递进",
    "核心问题": "如何在分心的环境中实现高质量产出？",
    # Step 2: Interpretive

    "关键术语": [
        {"term": "深度工作", "author_def": "在无干扰的状态下进行的专业活动", "diff": "与普通的忙碌不同"},
    ],
    "核心命题": ["深度工作能显著提升个人价值", "浅层工作导致职业停滞"],
    "论证链": "作者引用心理学研究与案例证明深度工作提升绩效。",
    # Step 3: Critical

    "时代局限": ["2010‑s 社交媒体兴起后注意力稀缺"],
    "立场盲点": ["作者为学术界工作者，未充分考虑创意行业的不同需求"],
    "未被证明的假设": ["所有职业都能受益于深度工作"],
    "最强反对意见": ["对快速迭代的行业，浅层工作可能更有效"],
    # Step 4: Applicability

    "可_skill化内容": ["时间块规划框架", "深度工作仪式"],
    "不适合_skill化内容": ["个人故事"],
    "预估_skill数量": 4,
    "优先级排序": ["时间块规划框架", "深度工作仪式"],
}

# Render and persist the Stage 0 overview

rendered = template.render(**context)
output_path = Path("./books/deep-work/BOOK_OVERVIEW.md")
output_path.parent.mkdir(parents=True, exist_ok_ok=True)
output_path.write_text(rendered, encoding="utf-8")
print(f"Stage 0 Adler analysis written to {output_path}")

The actual pipeline uses repository extractor scripts to populate this context automatically rather than manual entry.

Relationship to Downstream Stages

According to methodology/00-overview.md, the Stage 0 Adler analysis output is referenced by:

• Stage 1 extractors — which mine specific skills using the BOOK_OVERVIEW.md as grounding context
• The Boundary (B) field of every generated skill — populated from the Critical analysis to flag limitations

Without completing Adler analysis, subsequent stages lack the authoritative source context needed to generate valid, bounded skills.

Summary

• Adler analysis in Stage 0 is mandatory whole‑book comprehension before any skill extraction in cangjie‑skill
• The four steps—Structural, Interpretive, Critical, Applicability—mirror and extend the classic Adler reading method
• Output is a rendered BOOK_OVERVIEW.md from templates/BOOK_OVERVIEW.md.template that provides global context for all downstream stages
• The Critical step directly feeds skill Boundaries, preventing overgeneralization of extracted techniques

Frequently Asked Questions

Why is it called "Adler" analysis?

The name references the Adler reading‑comprehension method, specifically its three‑step cycle of Structural → Interpretive → Critical reading. Cangjie‑skill adds a fourth Applicability step to bridge comprehension toward actionable skill extraction, formalizing the entire process into a reproducible pipeline.

What happens if I skip Stage 0 Adler analysis?

Skipping Stage 0 means downstream extractors operate without global source context. Per SKILL.md configuration, this leads to skills missing proper Boundary (B) fields, resulting in overgeneralized or decontextualized outputs that may misapply the author's techniques to unsuitable situations.

How long should Adler analysis take relative to the full pipeline?

The methodology/01-stage0-adler.md documentation implies Stage 0 is deliberately time‑intensive—requiring complete source ingestion—because it amortizes across all downstream skills. A thorough Stage 0 prevents costly re‑work in later stages when boundary conditions or core arguments are discovered to be misunderstood.

Can Adler analysis handle non‑book sources like videos or podcasts?

Yes. The 结构_类型 (work type) field accepts any source classification, and the four analysis steps are media‑agnostic. The repository's template structure accommodates books, videos, podcasts, or courses by adjusting which structural elements are emphasized during Step 1 analysis.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}