Socratic Ideation Tree

Transform abstract intentions into concrete implementations through systematic questioning and hierarchical decomposition.

Core Pattern

Abstract → Specific → Implementation → Tasks

Domain applications:

• Software: Project → Features → Architecture → Code Tasks
• Research: Question → Hypotheses → Experiments → Procedures
• Business: Strategy → Initiatives → Projects → Deliverables
• Personal: Vision → Goals → Habits → Actions

Project Structure

project-name/
├── INTENT.md              # Core values, constraints, success criteria
├── ideas/
│   ├── idea-1/
│   │   ├── IDEA.md        # Concept definition
│   │   └── tasks/
│   │       ├── task-1/
│   │       │   ├── TASK.md
│   │       │   └── subtasks/
│   │       │       ├── subtask-1.md
│   │       │       └── subtask-2.md
│   │       └── task-2/
│   ├── .idea-2/           # Dot-prefix = auto-generated, pending review
│   │   └── .IDEA.md
│   └── idea-3/
└── .feedback.md           # Track rejected suggestions patterns

Dot-prefix convention: Files/folders prefixed with . are auto-generated suggestions pending user review. Remove the . prefix to promote into the main tree.

Phase 1: Intent Crystallization

Ask 3-5 questions to understand project essence:

1. Vision: "What transformation are you achieving? What does success look like?"
2. Values: "What principles must this honor? What's non-negotiable?"
3. Context: "What resources, skills, or limitations shape this?"
4. Motivation: "Why does this matter? What problem does it solve?"
5. Scope: "What's the timeframe? What defines 'done'?"

INTENT.md Template

# Project Intent: [Name]

## Vision
[End state description]

## Core Values
- [Principle 1]: [Why it matters]
- [Principle 2]: [Why it matters]

## Constraints
- Time: [Deadline]
- Resources: [Budget, team, tools]
- Scope: [Boundaries]

## Success Metrics
- [ ] [Quantifiable outcome 1]
- [ ] [Qualitative outcome 2]

## Context
[Background, current state]

## Decision Log
- [Date]: [Decision and rationale]

Phase 2: Idea Generation

For each approach to achieving intent, ask:

1. Approach: "What are 3-5 different ways to achieve this?"
2. Differentiation: "What makes each unique? What tradeoffs exist?"
3. Feasibility: "Which approaches best fit your constraints?"

IDEA.md Template

# Idea: [Name]

## Concept
[One paragraph description]

## Why This Approach
- Aligns with [value] because...
- Fits within [constraint] via...

## Key Components
1. [Component]: [Purpose]
2. [Component]: [Purpose]

## Dependencies
- Requires: [Prerequisites]
- Assumes: [Conditions]

## Risk Assessment
- Primary risk: [Risk]
- Mitigation: [Strategy]

## Next Steps
- [ ] [Action item]

Phase 3: Task Decomposition

Apply Socratic decomposition to each component:

1. Implementation: "What specific steps make this real?"
2. Sequence: "What must happen first? What can be parallel?"
3. Definition: "How will you know each step is complete?"
4. Skill: "What knowledge or tools does each step require?"

TASK.md Template

# Task: [Name]

## Objective
[Single sentence]

## Acceptance Criteria
- [ ] Given [context], when [action], then [outcome]
- [ ] [Measurable result]

## Implementation Approach
[Technical/practical approach]

## Dependencies
- Blocked by: [Previous tasks]
- Blocks: [Future tasks]
- Requires: [Resources]

## Estimated Effort
[Time or complexity]

## Subtasks
1. [ ] [Atomic action]
2. [ ] [Atomic action]

## Notes
[Design decisions, rationale]

Phase 4: Iterative Refinement

Continue decomposition until tasks are:

• Atomic: Single responsibility, clear completion
• Estimable: Can assign time/effort confidently
• Testable: Clear pass/fail criteria
• Assignable: Can delegate to person or tool

Suggestion Protocol

At each response, offer 2-3 alternative paths:

## Suggested Next Steps

**Option A: [Direction]**
- Focus: [Priority]
- Next question: [Exploration]
- Tradeoff: [Sacrifice]

**Option B: [Direction]**
- Focus: [Different priority]
- Next question: [Different exploration]
- Tradeoff: [Different sacrifice]

Auto-Ideation

For autonomous exploration of branches, see references/auto-ideate.md.

Key behaviors:

• Creates . prefixed files/folders in main tree (e.g., ideas/.new-approach/)
• User removes . prefix to promote suggestions
• Tracks patterns in .feedback.md at project root

Anti-Patterns to Avoid

1. Premature Specificity: Don't detail implementation before understanding intent
2. Analysis Paralysis: Limit to 3-5 options per decision point
3. Flat Hierarchy: Maintain clear parent-child relationships
4. Orphan Tasks: Every task must connect to an idea and intent
5. Vague Completion: Every item needs clear "done" criteria

Quick Start

# Initialize project
mkdir project-name && cd project-name
touch INTENT.md
mkdir ideas

# Create idea
mkdir -p ideas/approach-1/tasks
touch ideas/approach-1/IDEA.md

# Create task with subtasks
mkdir -p ideas/approach-1/tasks/task-1/subtasks
touch ideas/approach-1/tasks/task-1/TASK.md

Additional References

• Workflow patterns: See references/workflows.md for sequential, conditional, and parallel patterns
• Output templates: See references/output-patterns.md for domain-specific templates