Flow Diagrams
Five maps. Each one answers a question. Each answer feeds the next question. By the end, the PromptDeck is named as an instrument in the system language.
1 of 5
What does success look like?
OUTCOME MAP
═══════════════════════════════════════════════════════════════════════
A person compresses their world into five priorities on one page.
The empty cells become their strategy.
They return weekly to update it.
│
├── What exists ├── What blocks
│ ├── CDD prototype (5 patterns) │ ├── Zero external users
│ ├── Figma template (visual done) │ ├── "What if I pick the wrong five?"
│ ├── 1,200+ pages prove the pattern │ ├── People fill once, never return
│ ├── Stackmates auth │ ├── Identity & Access blocked
│ └── Convex real-time data layer │ └── "Self-reflection feels cringe"
│
└── Success measures (binary)
├── ≥ 10 external users fill a grid ─── YES / NO
├── D7 return rate > 35% ─────────── YES / NO
├── Grid completion > 60% cells ──── YES / NO
└── ≥ 3 grids shared externally ──── YES / NO
═══════════════════════════════════════════════════════════════════════Success is defined. Where does the current process fail? ↓
2 of 5
Where does time die?
VALUE STREAM MAP
═══════════════════════════════════════════════════════════════════════
[Too many priorities — new venture, quarter, crisis]
│
▼
1. CAPTURE 60 min wait: 0 50% complete & accurate
Write everything ─────────────────────────────────────────────────
│
▼
2. COMPRESS 30 min wait: 1-3 DAYS 30% complete & accurate
Pick what matters. ─────────░░░░░░░░░░░░░░──────────────────────────
Energy drops. energy drops, no structure forces completion
│
▼
3. DOCUMENT 3 hrs wait: 1-3 DAYS 20% complete & accurate
Slides, plan. ─────────░░░░░░░░░░░░░░──────────────────────────
Blank page. procrastination, blank page problem
│
▼
4. SHARE 30 min wait: 3-7 DAYS 40% complete & accurate
Present to team ─────────░░░░░░░░░░░░░░░░░░░░░░░░───────────────
scheduling, format mismatch
│
▼
5. UPDATE never wait: ∞ 0% complete & accurate
Always fails. ──────────────────────── STEP 5 NEVER HAPPENS ──
│
▼
[Frozen artifact. Clarity decays. Repeat next quarter.]
┌─────────────────────────────────────────────────────────────────┐
│ Cycle time: ~5 hours Flow efficiency: ~2% │
│ Wait time: ~11 days World-class: >40% │
│ Lead time: ~16 days THE WASTE ISN'T CREATION. │
│ IT'S DECAY. │
└─────────────────────────────────────────────────────────────────┘
═══════════════════════════════════════════════════════════════════════2% flow efficiency. Step 5 never happens. The instrument fixes Step 5.
The bottleneck is clear. What blocks the fix? ↓
3 of 5
What blocks the fix?
DEPENDENCY MAP
═══════════════════════════════════════════════════════════════════════
START
│
┌────────────────┼────────────────┐
│ │ │
▼ ▼ ▼
┌──────────────┐ ┌──────────────┐ ┌──────────────┐
│ Identity & │ │ CDD Proto- │ │ Figma │
│ Access │ │ type │ │ Template │
│ ═══ HARD │ │ (exists) │ │ (exists) │
│ BLOCKED │ └──────┬───────┘ └──────┬───────┘
└──────┬───────┘ │ │
│ └────────┬────────┘
▼ │
┌──────────────┐ │
│ Auth Wiring │ │
│ 2 days │ │
└──────┬───────┘ │
│ │
▼ ▼
┌──────────────┐ ┌──────────────┐
│ Convex │ │ Grid Engine │
│ Persistence │ │ 3 days │
│ ═══ HARD │ │ (from proto) │
│ 2 days │ └──────┬───────┘
└──────┬───────┘ │
│ │
└────────┬────────────────┘
│
▼
┌──────────────┐
│ Cell Editor │
│ 3 days │
└──────┬───────┘
│
▼
┌──────────────┐
│ Mode Toggle │
│ 2 days │
└──────┬───────┘
│
┌───────────┴───────────┐
▼ ▼
Sprint 1 (5d) Sprint 2 (5d)
Card + Scale + Snapshots +
Highlighting Templates + Share
│ │
└───────────┬───────────┘
▼
User Testing
5 pilot users
═══ = HARD (cannot proceed without)
─── = Soft (can proceed with mocks or exists)
Critical path: ~22 engineering days + Identity fix wait
═══════════════════════════════════════════════════════════════════════Two hard blockers. Everything else exists or is soft. What can we build with? ↓
4 of 5
What can we build with?
CAPABILITY MAP
═══════════════════════════════════════════════════════════════════════
READY (maturity 3+) TO BUILD (maturity 0)
────────────────── ─────────────────────
Domain Content (1,200+ pages) █████ 5 Cell Editor ░░░░░ P1
Email Delivery (Resend) █████ 5 Convex Persistence ░░░░░ P1
Figma Template ████░ 4 Mode Toggle ░░░░░ P2
Stackmates Auth ████░ 4 Card View (Stack) ░░░░░ P2
CDD Prototype (5 patterns) ███░░ 3 Scale Selector ░░░░░ P2
Convex DB ███░░ 3 Empty Cell Highlight ░░░░░ P2
Grid UI (Mission Control) ███░░ 3 Snapshot System ░░░░░ P2
Template Engine ░░░░░ P3
7 capabilities ready Share / Export ░░░░░ P3
AI Coaching ░░░░░ P3
10 capabilities to build
┌─────────────────────────────────────────────────────────────────┐
│ 60% composition (wire existing pieces) │
│ 40% new code (persistence + production UI) │
│ The sprint is wiring, not invention. │
└─────────────────────────────────────────────────────────────────┘
═══════════════════════════════════════════════════════════════════════The pieces exist. How does the instrument bring them to life? ↓
5 of 5
How does the instrument work?
The PromptDeck is a navigation instrument. The A&ID names it in the system language.
AGENT & INSTRUMENT DIAGRAM
═══════════════════════════════════════════════════════════════════════
AGENTS INSTRUMENTS
(who applies force) (what verifies and rewards)
────────────────── ───────────────────────────
User (Human) Gap Highlighter
fills, reflects, returns weekly empty cells = strategy
Deck Engine (System) Snapshot Comparator
persists, highlights, snapshots this week vs last week
AI Coach (V3, future) Mode Lens
suggests questions for gaps pep talk ↔ pitch
THE FILL LOOP
─────────────────────────────────────────────────────────────────
FILL ──→ GAPS ──→ QUESTIONS ──→ RESEARCH ──→ UPDATE ──→ COMPARE
│ │ │ │ │
│ Gap Highlighter Domain Deck Snapshot
│ shows what's Content Engine Comparator
│ missing │ │
└──────────────────────── FEEDBACK ←───────────┴───────────┘
grid sharpens each pass
FOUR LOOPS, FOUR FREQUENCIES
─────────────────────────────────────────────────────────────────
SESSION Fill → see gap → fill next → complete instant
WEEKLY Return → snapshot → compare → update minutes
MODE Pep talk → pitch → see misalignment → refine minutes
COMPOUND Many snapshots → patterns → deeper five days
HORIZON
─────────────────────────────────────────────────────────────────
H1 Prompt Deck ──→ WHAT matters, WHERE gaps are
│
H2 Time+Mind ──→ closes the gap between WHAT and DID
│
H3 AI Coach ──→ fills what you can't see alone
═══════════════════════════════════════════════════════════════════════Five maps. The outcome names success. The value stream exposes the waste. The dependencies show what blocks. The capabilities show what exists. The A&ID names the instrument that fixes it all.
Questions
If the diagram doesn't make the engineer say "I can build this" — is the diagram wrong or the idea?
- Which map revealed something the PromptDeck slides didn't say?
- When the value stream shows 2% flow efficiency, does that validate the pain or expose a gap in the solution?
- Does the A&ID name the PromptDeck as an instrument that compounds — or just a tool that stores?