refactor(topics): 멀티 에이전트용 지식 재편 — _Common(공통 기본기) + Domain_* 구조

에이전트 8종(대화형/프로그래머 C·S/디자이너/설계자/기획자/QA/PD/PM)에게 [공통 기본 능력 + 롤별 Specialty] 2층으로 지식을 주입하기 위한 재분류. 문서 내용·포맷은 무수정, 폴더 이동만 (6,372개 문서 수 보존 확인). - Topic_Programming → Domain_Programming (내부 구조 보존) - Topic_Graphic → Domain_Design - Topic_Business → Domain_Product - Topic_General → Domain_General - _Common 신설: Math(구 Topic_Math_Specialty), Reasoning(구 General/From_Thinking & Reasoning), Reasoning_Creativity(구 General/From_창의성), Communication(Poetic_Blog_Writing + From_writing) - 타 도메인의 From_* 폴더는 유지 (출처 표기일 뿐, 이미 도메인에 맞게 분류된 문서) - 빈 폴더 정리 (memory/procedures) - 에이전트→폴더 매핑은 workspace의 .astra/agent-knowledge-map.json (9개 에이전트) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-11 11:05:56 +09:00
parent 6549ead309
commit c24165b8bc
6193 changed files with 1717 additions and 31 deletions
@@ -0,0 +1,166 @@
+---
+id: wiki-2026-0508-mece-pyramid-principle
+title: MECE + Pyramid Principle
+category: 10_Wiki/Topics
+status: verified
+canonical_id: self
+aliases: [MECE, Pyramid Principle, 미씨 + 피라미드, McKinsey Framework]
+duplicate_of: none
+source_trust_level: A
+confidence_score: 0.9
+verification_status: applied
+tags: [problem-solving, communication, structure, consulting, writing]
+raw_sources: []
+last_reinforced: 2026-05-10
+github_commit: pending
+tech_stack:
+  language: methodology
+  framework: mckinsey
+---
+
+# MECE + Pyramid Principle
+
+## 매 한 줄
+> **"매 MECE 는 thinking, Pyramid 는 communication"**. MECE (Mutually Exclusive, Collectively Exhaustive) 는 문제 분해 원칙, Pyramid Principle 은 결론-우선 communication structure. Barbara Minto (1973) 의 McKinsey 표준. 2026 LLM 시대에도 prompt structuring / report writing 의 backbone.
+
+## 매 핵심
+
+### 매 MECE
+- **Mutually Exclusive**: 각 카테고리 겹침 없음.
+- **Collectively Exhaustive**: 모든 가능성 포함.
+- 2x2 matrix, decision tree, issue tree 의 기본.
+
+### 매 Pyramid Principle
+- **Top**: governing thought / answer first.
+- **Middle**: 3-5 supporting arguments (MECE).
+- **Bottom**: data, evidence, examples.
+- **SCQA opener**: Situation → Complication → Question → Answer.
+
+### 매 응용
+1. Consulting deliverable / executive summary.
+2. Research paper structure.
+3. LLM prompt design (system + sections).
+4. Code review write-up.
+
+## 💻 패턴
+
+### Issue tree decomposition
+```python
+class Node:
+    def __init__(self, q, children=None):
+        self.q = q
+        self.children = children or []
+
+# Profit decline 분석
+tree = Node("Why is profit declining?", [
+    Node("Revenue down?", [
+        Node("Volume down?"),
+        Node("Price down?"),
+    ]),
+    Node("Cost up?", [
+        Node("COGS up?"),
+        Node("OpEx up?"),
+    ]),
+])
+# 매 each level MECE
+```
+
+### MECE validator
+```python
+def is_mece(categories: list[set]) -> tuple[bool, bool]:
+    universe = set.union(*categories)
+    # ME: pairwise disjoint
+    me = all(not (a & b) for i, a in enumerate(categories)
+                          for b in categories[i+1:])
+    # CE: union covers universe
+    ce = set.union(*categories) == universe
+    return me, ce
+```
+
+### Pyramid outliner (LLM)
+```python
+def pyramid_outline(question: str) -> dict:
+    prompt = f"""Structure as Pyramid Principle:
+1. Governing answer (1 sentence).
+2. 3 MECE supporting arguments.
+3. For each, 2-3 evidence bullets.
+
+Question: {question}
+Output JSON."""
+    resp = client.messages.create(
+        model="claude-opus-4-7",
+        max_tokens=2048,
+        messages=[{"role": "user", "content": prompt}],
+    )
+    return json.loads(resp.content[0].text)
+```
+
+### SCQA opener generator
+```python
+def scqa(situation, complication, question, answer):
+    return (
+        f"**Situation**: {situation}\n"
+        f"**Complication**: {complication}\n"
+        f"**Question**: {question}\n"
+        f"**Answer**: {answer}"
+    )
+```
+
+### 2x2 framework
+```python
+def matrix_2x2(items, axis_x, axis_y):
+    quadrants = {"high-high": [], "high-low": [],
+                 "low-high": [], "low-low": []}
+    for item in items:
+        x = "high" if axis_x(item) else "low"
+        y = "high" if axis_y(item) else "low"
+        quadrants[f"{x}-{y}"].append(item)
+    return quadrants
+```
+
+### Top-down report builder
+```python
+def build_report(answer, args: list[dict]):
+    out = [f"# {answer}\n"]
+    for i, arg in enumerate(args, 1):
+        out.append(f"## {i}. {arg['claim']}")
+        for ev in arg["evidence"]:
+            out.append(f"- {ev}")
+    return "\n".join(out)
+```
+
+## 매 결정 기준
+| 상황 | Approach |
+|---|---|
+| Problem decomposition | Issue tree (MECE at each level) |
+| Executive deck | Pyramid + SCQA + 3 args |
+| Categorization | 2x2 matrix or MECE list |
+| LLM task | Pyramid in system prompt |
+
+**기본값**: Issue tree 분석 → Pyramid 로 communicate.
+
+## 🔗 Graph
+- 부모: [[Problem Solving Process]]
+- 변형: [[Pyramid Principle]] · [[Issue Tree]]
+- 응용: [[Technical Writing]]
+- Adjacent: [[Hypothesis-Driven]]
+
+## 🤖 LLM 활용
+**언제**: report drafting, prompt structuring, decomposition assistance.
+**언제 X**: creative / divergent ideation — 매 over-constrains.
+
+## ❌ 안티패턴
+- **False MECE**: overlap 있는데 disjoint 라 가정.
+- **Bottom-up dump**: data 먼저 늘어놓고 conclusion 마지막 → executive 가 lost.
+- **Over-decomposition**: 7+ branches at one level → cognitive overload.
+- **Forced 3 categories**: 매 항상 3 으로 강제 → exhaustiveness 깨짐.
+
+## 🧪 검증 / 중복
+- Verified (Minto 1973 "The Pyramid Principle", McKinsey training docs, HBR 2019).
+- 신뢰도 A.
+
+## 🕓 Changelog
+| 날짜 | 변경 |
+|---|---|
+| 2026-05-08 | Phase 1 |
+| 2026-05-10 | Manual cleanup — issue tree + SCQA + 2x2 패턴 |