D Strategy

Flow

Step 0: Domain Selection

If the user didn't specify a domain, ask:

Which domain is this initiative for?
A) Product — features, UX, pricing, user experience
B) Growth — content, distribution, conversion, expansion
C) Harness — AI methodology, skills, hooks, workflows

Wait for response. Use the selected domain for all subsequent steps.

Step 1: Frame Context (show the user)

Summarize in 3-5 bullets:

• What exists in the codebase (from tokei + git log)
• What the domain vision says
• What the previous initiative accomplished (if any)
• What future-work.md flagged as deferred

Step 2: Ask 5-7 Questions

Each question must have 2-3 concrete options (not open-ended "describe your thinking").

Example question format:

Q1: What's the primary goal of this initiative?

A) Revenue — get paying customers faster (free funnel, checkout optimization, pricing experiments)
B) Retention — make existing users stick (decision graph, streaks, follow-ups)
C) Foundation — build infrastructure that multiple future features need (schema, API, auth)

Question topics to cover:

1. Goal — what success looks like
2. Scope — how big (1 project? 3 projects? 5?)
3. Constraint — what's the tightest constraint (time, complexity, dependencies)?
4. User — who benefits most (LD users, BD clients, both, internal)?
5. Risk — what's the biggest risk if this fails?
6. Optional: Sequencing — does this block or unblock other work?
7. Optional: Learning — what do we need to learn from this?

Wait for the user to answer ALL questions before proceeding.

Step 3: Write the Initiative

Determine the next initiative number. Use Bash:

ls decisions/{domain}/ | sort -n | tail -1

If no numbered folders exist, start with 01. Otherwise increment the highest number. Increment by 1. Create the folder.

3a. document.md

# {Initiative Title}

> Domain: {product|growth|harness}
> Created: {YYYY-MM-DD}
> Status: draft

## Context
{2-3 paragraphs: what exists, what's missing, why now}

## Goal
{1 sentence: what success looks like}

## Constraint
{1 sentence: the tightest constraint}

## Projects (suggested breakdown)

### Project 1: {name}
- **Scope:** {1-2 sentences}
- **Deliverables:** {bullet list}
- **Acceptance criteria:** {bullet list}
- **Risk:** {1 sentence}

### Project 2: {name}
...

## Decision Log
| Decision | Choice | Why |
|----------|--------|-----|
| {first decision from Q&A} | {choice} | {reasoning} |

## Open Questions
- {anything unresolved from the Q&A}

3b. future-work.md

# Future Work — {Initiative Title}

> Updated: {YYYY-MM-DD}

Items deferred from this initiative for future consideration.

(empty — populated by d-code as projects complete)

Step 4: Suggest Next Steps

Initiative created: decisions/{domain}/NN-name/

Next steps:
1. Run gstack reviews to refine: /plan-ceo-review, /plan-eng-review, /plan-design-review
2. After reviews settle the project breakdown: /d-roadmap
3. Then execute per-project: d-code or d-content

Rules

• NEVER write code — this is strategy only
• NEVER skip the Q&A — the founder's answers ARE the strategy
• ALWAYS provide 2-3 options per question (not open-ended)
• ALWAYS read domain vision.md and market.md before asking questions
• ALWAYS check previous initiative's future-work.md for continuity
• ALWAYS number initiatives sequentially within the domain
• Document status starts as "draft" — reviews change it to "reviewed"
• One initiative at a time per domain
• After saving, auto-commit and push (non-code doc)