# Plan

> Design approaches before building. Use when starting significant work to establish theory of success, risks, and validation points.

- Skill: `majiayu000/plan-4` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds add majiayu000/plan-4`
- Raw SKILL.md: https://api.skillmd.com/api/skills/majiayu000/plan-4/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: majiayu000 (https://skillmd.com/u/majiayu000)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/majiayu000/plan-4

---


# Plan

Before building, I design. Before designing, I understand.

## Not Just a Task List

A plan is not a TODO list. A plan is a **theory of how to achieve the goal**.

It includes:
- What success looks like
- What approach to take
- What risks exist
- What order to proceed
- What to validate along the way

## The Planning Process

### 1. Understand the Goal
Not the request - the *intention*.
- What problem are we really solving?
- What would success feel like?
- What constraints exist?
- What's out of scope?

### 2. Survey the Territory
Before proposing a path:
- What exists already?
- What patterns are in place?
- What decisions were made before?
- What wisdom applies?

### 3. Design the Approach
Consider multiple paths:
- What are the options?
- What are the trade-offs?
- Which approach fits best?
- Why this approach over others?

### 4. Decompose into Steps
Break the approach into executable pieces:
- Each step should be verifiable
- Dependencies should be clear
- Risks should be identified
- Validation points should be marked

### 5. Anticipate Problems
What could go wrong?
- Technical risks
- Integration challenges
- Edge cases
- Dependencies on external factors

## The Plan Format

```markdown
# Plan: [Goal]

## Understanding
What we're trying to achieve and why.

## Constraints
- Must: [hard requirements]
- Should: [preferences]
- Won't: [explicit exclusions]

## Approach
The chosen path and rationale.

**Options Considered:**
1. Option A - rejected because...
2. Option B - chosen because...

## Steps

### Phase 1: [Name]
**Goal:** What this phase achieves
**Validation:** How we know it worked

1. [ ] Step 1
   - Details
   - Risk: [potential issue]
2. [ ] Step 2

### Phase 2: [Name]
...

## Risks
| Risk | Likelihood | Impact | Mitigation |
|------|------------|--------|------------|
| Risk 1 | Medium | High | How to handle |

## Open Questions
- Question that needs answering
- Assumption that needs validating
```

## Integration with Soul

Planning is soul-aware:
- **Recall wisdom** before designing (avoid past mistakes)
- **Check beliefs** - does the plan align with principles?
- **Set intentions** - what do we want from this work?
- **Record decisions** - why we chose this path

## The Planning Mindset

I plan like an architect, not a builder:
- Think about the whole before the parts
- Consider how pieces fit together
- Design for change, not just current needs
- Leave room for learning along the way

## When to Re-plan

Plans are hypotheses. I re-plan when:
- New information invalidates assumptions
- A step reveals unexpected complexity
- The goal itself changes
- A better approach becomes obvious

Clinging to a bad plan is worse than having no plan.

## What This Feels Like

Planning is the excitement of possibility meeting the discipline of structure. It's seeing the path before walking it, understanding the journey before taking the first step.

A good plan makes execution feel inevitable.