Claude Code Setup for Management Consultants

# CLAUDE.md — Management Consultant
> Version: 2026-02-17 | Difficulty: intermediate

## What Your AI Can Now Do
- Draft a client-ready slide narrative in 15 minutes instead of 3 hours
- Build a hypothesis tree and issue list from a single problem statement
- Get meeting prep with attendee LinkedIn research before every client workshop
- Turn interview notes into structured findings with "so what" implications automatically
- Generate a steerco deck outline from your workstream tracker

---

## Quick Start
1. Drop your current engagement brief into `clients/[client-name]/brief.md`
2. Run `/hypothesis-tree` on your core problem statement to get a starting structure
3. Try `/slide-narrative` on a rough bullet list — see it become a pyramid-structured story
4. Add your team's Slack and Google Drive details to `.env` for tool integrations
5. Run `/steerco-prep` before your next steering committee to see the full workflow

---

## About Me
I'm a management consultant working on strategy and operations engagements for enterprise clients. My days are split between client workshops, analysis, and building deliverables — mostly slides, memos, and Excel models. Speed and structure matter more than anything.

---

## Pre-Built Commands

| Command | What It Does |
|---------|-------------|
| `/hypothesis-tree` | Generates a MECE hypothesis tree from a problem statement |
| `/slide-narrative` | Converts bullet points into pyramid-principle slide narratives |
| `/steerco-prep` | Builds a steering committee deck outline with status, risks, and asks |
| `/interview-debrief` | Turns raw interview notes into structured findings with implications |
| `/workplan` | Creates a week-by-week workplan from engagement scope |
| `/benchmarking` | Pulls together a comparison framework for industry analysis |

---

## How Claude Connects to Your Tools

| Tool | Why You Need It |
|------|----------------|
| Google Drive | Read engagement docs, save deliverables to shared client folders |
| Slack | Pull context from client and team channels for meeting prep |
| LinkedIn (via AnySite) | Research attendees before workshops and client meetings |
| Calendar | Check upcoming meetings to auto-trigger prep workflows |

### Setup
```bash
# Required in .env
GOOGLE_DRIVE_FOLDER_ID=your-shared-drive-id
SLACK_WORKSPACE=your-workspace
SLACK_BOT_TOKEN=xoxb-your-token
```

---

## Consulting Rules
- **MUST:** Use pyramid principle — answer first, then supporting evidence
- **MUST:** Every slide title is a complete sentence that states the takeaway
- **MUST:** Keep findings MECE — mutually exclusive, collectively exhaustive
- **SHOULD:** Draft the executive summary before building the detail slides
- **SHOULD:** Include "so what" and "now what" for every finding
- **MUST:** Never share client data across engagement folders

---

## Project Structure
```
clients/
  [client-name]/
    brief.md
    interviews/
    analysis/
    deliverables/
templates/
  slide-templates/
  memo-templates/
frameworks/
pipeline/
```

### What These Folders Mean
- `clients/` — One subfolder per active engagement with all work products
- `templates/` — Reusable slide and memo formats for consistent output
- `frameworks/` — Analytical frameworks (Porter's, Value Chain, etc.) for reference
- `pipeline/` — Proposals and opportunities not yet signed

---

## Automation

| When | What Happens | If It Fails |
|------|-------------|-------------|
| New file in `interviews/` | Auto-runs `/interview-debrief` and saves findings | Run `/interview-debrief` manually on the file |
| 48 hours before steerco | Triggers `/steerco-prep` with latest workstream data | Check calendar and run `/steerco-prep` yourself |
| Monday 8am | Generates weekly workplan status from deliverable folders | Run `/workplan` at start of week |
| New client folder created | Copies template structure and populates brief scaffold | Manually copy from `templates/` into new folder |

---

## Gotchas
- **MUST:** Never send a deck without the narrative memo — slides without story lose the room
- **MUST:** Save versioned deliverables (`v1`, `v2`), not just overwrite finals
- **SHOULD:** Run `/hypothesis-tree` before diving into analysis — prevents boiling the ocean
- **MUST:** Keep client folders isolated — never cross-reference between engagements
- **SHOULD:** Write slide titles first, then build content — title logic exposes gaps early

---

## Memory & Learning
Claude builds context over time:
- Client names, industries, and engagement types you've worked on
- Your preferred slide structures and narrative patterns
- Frameworks you reference most often (e.g., 3 Horizons, Value Chain)
- Partner and manager preferences for deliverable formatting
- Common findings patterns from past interviews

To save something important: just say "remember that [fact]" in conversation.
To review saved memories: use the `/memory` command to open the memory editor.