Files

T

Antigravity Agent 9148c358d0 docs(10_Wiki): 위키 전체 재구성 — Topic_* 폴더를 4개 카테고리로 통합 + 대규모 중복 제거

Topic_Agent/Topic_Blog/Topics/Topics_Biz/Topics_Meeting/Topics_Rag의 마크다운 지식 문서를
Topic_General/Topic_Programming/Topic_Graphic/Topic_Business 4개 카테고리로 재분류.

- 중복 제거: frontmatter의 status:duplicate/merged + duplicate_of/redirect_to 필드로
  자기 자신을 중복으로 선언한 리다이렉트 stub 1032개 제거, 완전 동일 내용 파일 472개 제거,
  동일 파일명·다른 내용 충돌 시 더 큰(완전한) 버전만 유지(162개 제거) — 총 1639개 중복 제거.
- 분류: 폴더 단위로 명확한 항목(AI_and_ML/Coding/Architecture 등 → Programming,
  Comfyui/Visual_Effects → Graphic, Topics_Biz/Topics_Meeting/사업 등 → Business,
  Poetic_Blog_Writing/창의성/Game_Design 등 → General)은 폴더 우선순위로,
  나머지 혼재 폴더(Topic_Agent/Topic_Blog/Topics 루트/Thinking & Reasoning/Other/UI_UX_Assets)는
  title/tags 키워드 스코어링으로 파일 단위 분류(불명확한 경우 General로 폴백).
  원본 폴더명은 "From_*" 서브폴더로 보존해 추적 가능성 유지.
- 최종 배치: Programming 2784 / General 1608 / Graphic 285 / Business 249 = 4926개 문서.
- 에이전트 운영 상태(.astra/.agent/.obsidian/sessions/memory/_company/docs/lessons/_shared/src)는
  지식 콘텐츠가 아니므로 재분류 대상에서 제외하고 원위치 유지.
- Topics/Topic_email(상위 보호 폴더 Topic_email과 파일명 100% 중복) 삭제 — 보호 폴더 자체는 미변경.
- 완전히 비게 된 Topic_Agent/Topic_Blog/Topics_Biz/Topics_Rag 폴더 제거.

2026-07-05 00:33:48 +09:00

5.3 KiB

Raw Blame History

id, title, category, status, canonical_id, aliases, duplicate_of, source_trust_level, confidence_score, verification_status, tags, raw_sources, last_reinforced, github_commit, tech_stack

title

Management Consulting

매 한 줄

"매 management consulting 은 hypothesis-driven problem solving as a service". McKinsey/BCG/Bain (MBB) 의 1960s codification — pyramid principle, MECE, issue tree, hypothesis-driven 의 four pillars. 매 2026 modern state: AI augmentation (Claude Opus 4.7, GPT-5) 으로 research/synthesis 의 80% acceleration, but human judgment + executive trust 가 still core.

매 핵심

매 four pillars

Pyramid principle (Minto): top answer first, supporting reasons next, evidence below.
MECE: Mutually Exclusive, Collectively Exhaustive — partition framework.
Issue tree: top question → sub-questions, recursively.
Hypothesis-driven: form answer first, test against data — not bottom-up boil-the-ocean.

매 typical engagement structure

Week 1-2: scoping, interviews, hypothesis tree.
Week 3-6: data gathering, model building, expert calls.
Week 7-9: synthesis, slide drafting, partner reviews.
Week 10-12: client workshops, final readout, implementation roadmap.

매 modern (2026) augmentation

AI research: Claude/GPT for industry primers, expert call prep, public filings synthesis.
AI modeling: code-interpreter for forecasts, sensitivity tables.
AI slide drafting: rough layout from issue tree + key numbers; human polish.
Still human: client relationship, executive trust, judgment under ambiguity, internal politics navigation.

매 firm tiers

MBB: McKinsey, BCG, Bain.
Tier 2: Strategy&, Oliver Wyman, LEK, Roland Berger, Kearney.
Big 4 strategy: Deloitte Monitor, EY-Parthenon, PwC Strategy&, KPMG.
Boutique: Veritas, Putnam, Analysis Group (specialized).

💻 패턴

Issue tree as data

interface IssueNode {
  question: string;
  hypothesis?: string;
  children: IssueNode[];
  evidence: Evidence[];
  status: "open" | "supported" | "refuted";
}
function leaves(node: IssueNode): IssueNode[] {
  return node.children.length === 0 ? [node] : node.children.flatMap(leaves);
}

MECE check

function isMECE<T>(partition: T[][], universe: Set<T>): { mutually: boolean; exhaustive: boolean } {
  const flat = partition.flat();
  const mutually = flat.length === new Set(flat).size;
  const exhaustive = [...universe].every((x) => flat.includes(x));
  return { mutually, exhaustive };
}

Pyramid principle slide skeleton

# [Action title: the answer in one sentence]
- Reason 1: [supporting argument]
  - Evidence A
  - Evidence B
- Reason 2: [supporting argument]
- Reason 3: [supporting argument]

Profitability tree (canonical)

Profit
├── Revenue
│   ├── Volume × Price
│   │   ├── Market size × Share
│   │   └── Mix × Discount
└── Cost
    ├── COGS (variable)
    └── SG&A (fixed)

Expert call synthesis prompt (Claude Opus 4.7)

const prompt = `
You are a research analyst. Given these 5 expert call transcripts on [TOPIC],
extract:
1. Areas of consensus (≥3 experts agree)
2. Areas of disagreement
3. Quantitative anchors (market size, growth, margin)
4. Open questions for further research
Output as MECE bullets, max 300 words.
`;

2x2 framework template

|              | High Impact | Low Impact |
|--------------|-------------|------------|
| Easy to do   | DO NOW      | Quick wins |
| Hard to do   | Strategic   | DROP       |

매 결정 기준

상황	Approach
C-suite strategy refresh	MBB or Tier 2 strategy boutique
Operational turnaround	Big 4 + ops specialists (AlixPartners)
M&A due diligence	Bain (PE focus), strategy boutiques
Digital/AI transformation	McKinsey QuantumBlack, BCG X, Bain Vector
In-house build	Hire ex-consultant + AI tooling

기본값: Hypothesis-driven + issue tree + MBB-style synthesis. AI augmentation for research/modeling. Human for trust/judgment.

🔗 Graph

🤖 LLM 활용

언제: industry primer, expert call prep, slide drafting, financial modeling, synthesis. 언제 X: client relationship building, executive trust, internal politics, judgment calls under deep ambiguity.

❌ 안티패턴

Boil the ocean: hypothesis 없이 모든 data 모음 → time/budget overrun.
Pretty slides, weak answer: aesthetics > insight 의 trap.
Recommendation without data: "we believe" without grounding.
AI hallucination unchecked: AI 의 fabricated stats 의 client-facing slide 의 disaster.

🧪 검증 / 중복

Verified (Minto's Pyramid Principle, McKinsey/BCG/Bain public materials, 2026 industry observation).
신뢰도 A.

🕓 Changelog

날짜	변경
2026-05-08	Phase 1
2026-05-10	Manual cleanup — FULL spec rewrite with 2026 AI augmentation context

5.3 KiB Raw Blame History Unescape Escape