[!WARNING] Experimental Feature / 實驗性功能

This feature is under active development and may change significantly in v4.0. 此功能正在積極開發中，可能在 v4.0 中有重大變更。

Methodology System Skill

Language: English | 繁體中文

Version: 2.0.0 Last Updated: 2026-01-25

Overview

The Methodology System provides a unified framework for managing development methodologies in projects that adopt Universal Development Standards.

Two Independent Systems

┌────────────────────────────────────────────────────────────────────────────┐
│                  Two Independent Methodology Systems                         │
├────────────────────────────────────────────────────────────────────────────┤
│                                                                            │
│  System A: SDD (AI-Era Methodology)                                        │
│  ─────────────────────────────────────                                     │
│  /spec → Review → /derive-all → Implementation → Verification              │
│  Best for: New projects, AI-assisted development, greenfield features     │
│                                                                            │
│  System B: Double-Loop TDD (Traditional)                                   │
│  ─────────────────────────────────────                                     │
│  /bdd (Outer Loop) → /tdd (Inner Loop) → Demo                              │
│  Best for: Legacy systems, manual development, established codebases      │
│                                                                            │
│  Optional Input: ATDD Workshop                                             │
│  ─────────────────────────────────────                                     │
│  Stakeholder collaboration that feeds into EITHER system                   │
│                                                                            │
└────────────────────────────────────────────────────────────────────────────┘

Supported Methodologies

System A: SDD (Spec-Driven Development)

• Proposal → Review → Forward Derivation → Implementation → Verification → Archive
• AI-native workflow with spec as authoritative source

System B: Double-Loop TDD

• BDD (Outer Loop) - Discovery → Formulation → Automation
• TDD (Inner Loop) - Red → Green → Refactor

Optional Input

• ATDD - Acceptance Test-Driven Development workshop (feeds into either system)

Custom - User-defined methodologies

Features

1. Phase-Aware Guidance

AI automatically tracks the current phase and provides context-appropriate guidance:

┌─────────────────────────────────────────────┐
│ 📋 Current Methodology: TDD                  │
│ 📍 Current Phase: 🔴 RED (1-5 min)           │
│                                             │
│ Checklist:                                  │
│   ✅ Test describes behavior                │
│   ✅ Test name is clear                     │
│   ⬜ Test follows AAA pattern               │
│   ⬜ Test fails when run                    │
│                                             │
│ Next: Write the test following AAA pattern  │
└─────────────────────────────────────────────┘

2. Checkpoint Reminders

Automatic reminders based on methodology triggers:

• Phase Transition: Suggest commit when phase completes
• Accumulation Warning: Warn when changes exceed threshold
• Skip Warning: Alert after consecutive skipped check-ins

3. Methodology Switching

Switch between methodologies as project needs change:

/methodology switch bdd

4. Custom Methodology Support

Define team-specific workflows in .standards/methodologies/:

id: my-team-workflow
name: My Team Workflow
phases:
  - id: plan
    name: Planning
    checklist:
      - id: requirements-clear
        text: Requirements understood
        required: true

Commands

Configuration

Methodology configuration is stored in .standards/manifest.json:

{
  "methodology": {
    "active": "tdd",
    "available": ["tdd", "bdd", "sdd", "atdd"],
    "config": {
      "tdd": {
        "checkpointsEnabled": true,
        "reminderIntensity": "suggest",
        "skipLimit": 3
      }
    }
  }
}

Configuration Options

AI Behavior

Detection

1. Check .standards/manifest.json for methodology.active
2. Load methodology definition from:
   • Built-in: methodologies/{id}.methodology.yaml
   • Custom: .standards/methodologies/{id}.methodology.yaml

Phase Tracking

• Track current phase based on trigger conditions
• Update phase when exit conditions are met
• Provide phase-specific guidance and checklist

Context Keywords

AI will automatically activate methodology context when these keywords are detected:

Integration with Other Standards

Check-in Standards

When a phase completes, the methodology system integrates with checkin-standards.md:

Phase GREEN completed.

Changes:
- Files: 3
- Lines: +45 / -2

Suggested commit:
  test(auth): add login validation test
  feat(auth): implement login validation

[1] Commit now  [2] Continue working  [3] View changes

Code Review

Additional review checks are added based on active methodology:

• TDD: Tests follow naming conventions, single behavior per test
• BDD: Declarative style, reusable steps
• SDD: Change matches spec, no scope creep
• ATDD: All acceptance criteria have tests

Related Skills

• TDD Assistant - Detailed TDD guidance
• Spec-Driven Dev - SDD workflow
• Code Review Assistant - Review integration

Files

• integrated-flow.md - Complete workflow guide for both systems
• runtime.md - AI behavior and runtime guide
• create-methodology.md - Custom methodology creation wizard

Version History