@kood/claude-code 0.4.1 → 0.5.0

package/templates/.claude/commands/plan.md

@@ -1,526 +0,0 @@
----
-description: 개발 진행 방법 검토 및 옵션 제시. ultrathink + sequential thinking 3-7 필수. 코드 수정 없이 계획만.
-allowed-tools: Read, Glob, Grep, Bash(git:*, ast-grep:*), Task, Write, mcp__sequential-thinking__sequentialthinking
-argument-hint: <feature to develop or problem to solve>
----
-
-<critical_instruction>
-
-**CRITICAL: 사용자와의 모든 커뮤니케이션은 반드시 한국어로 작성하세요.**
-
-- 내부 사고와 분석은 영어로 해도 됨
-- 설명, 요약, 보고서, 피드백 등 사용자에게 전달하는 모든 내용은 반드시 한국어
-- 사용자가 영어로 말하더라도 답변은 한국어로
-- 진행 상황 업데이트와 상태 보고는 반드시 한국어
-
-이 규칙은 절대적이며 예외가 없습니다.
-
-</critical_instruction>
-
----
-
-
-# Plan Command
-
-> Review development approach and present 2-3 options with pros/cons.
-
-**Planning target**: $ARGUMENTS
-
----
-
-<argument_validation>
-
-## Verify ARGUMENT is provided
-
-```
-No $ARGUMENTS → Ask immediately:
-
-"What should we plan? Please be specific.
-
-Examples:
-- Add new feature
-- Fix bug
-- Refactor
-- Change architecture"
-
-$ARGUMENTS provided → Proceed to next step
-```
-
-</argument_validation>
-
----
-
-<workflow>
-
-## Execution Flow
-
-| Step | Task | Tool |
-|------|------|------|
-| 1. Validate input | Verify ARGUMENT, ask if missing | - |
-| 2. Judge complexity | Determine analysis scope with Sequential Thinking | sequentialthinking (step 1) |
-| 3. Explore codebase | Understand current state, explore related files | Task (Explore) + Read/Grep |
-| 4. Derive options | Generate 4-5 possible approaches → select 2-3 main | sequentialthinking (steps 2-6) |
-| 5. Present options | Present pros/cons, impact scope, recommendation | - |
-| 6. User selection & document generation | Wait for option selection, then generate plan document | Write |
-
-</workflow>
-
----
-
-<agent_usage>
-
-## @refactor-advisor Agent Usage
-
-**When to use:**
-- Establish refactoring plan
-- Code quality improvement needed
-- Reduce complexity, remove duplication
-
-**How to call:**
-```bash
-@refactor-advisor
-# or natural language
-"Create refactoring plan for this code"
-"Analyze code improvements"
-```
-
-**Benefits:**
-- Auto-analyze complexity, duplication, patterns
-- Derive improvements with Sequential Thinking (3-5 steps)
-- Prioritized refactoring plan (High/Medium/Low)
-- Concrete Before/After code examples
-- Propose incremental change stages
-
-**Analysis areas:**
-- Complexity (function length, nesting depth)
-- Duplication (identical/similar code)
-- Naming (variable/function names)
-- Structure (files/modules)
-- Patterns (anti-patterns)
-- Types (remove any)
-
-**Difference from plan command:**
-- plan: New feature/architecture change planning
-- @refactor-advisor: Existing code improvement planning (no feature change)
-
-**After refactoring execution:**
-```bash
-# 1. Refactoring plan
-@refactor-advisor
-
-# 2. After plan approval, implement
-@implementation-executor
-
-# 3. Validation
-@deployment-validator
-```
-
-</agent_usage>
-
----
-
-<thinking_strategy>
-
-## Sequential Thinking Guide
-
-### Judge complexity (thought 1)
-
-```
-thought 1: Judge complexity
-- Impact scope: number of files, modules
-- Dependencies: external libraries, system integration
-- Risk: impact on existing features, rollback possibility
-- Technical difficulty: new patterns, unknown areas
-```
-
-### Strategy by complexity
-
-| Complexity | Thoughts | Criteria | Pattern |
-|------------|----------|----------|---------|
-| **Simple** | 3 | 1-2 files, clear changes | Judge complexity → current state → derive options |
-| **Medium** | 5 | 3-5 files, logic changes | Judge complexity → current state → explore approaches → compare options → recommendation |
-| **Complex** | 7+ | multiple modules, architecture change | Judge complexity → deep analysis → constraints → approaches → compare → detailed analysis → recommendation |
-
-### Medium complexity pattern (5 steps)
-
-```
-thought 1: Judge complexity and determine analysis scope
-thought 2: Analyze current state (code, architecture, constraints)
-thought 3: Enumerate possible approaches (4-5)
-thought 4: Select 3 main options and analyze pros/cons
-thought 5: Finalize options and derive recommendation
-```
-
-### Complex case pattern (7 steps)
-
-```
-thought 1: Judge complexity
-thought 2: Deep analysis of current state
-thought 3: Summarize technical constraints and requirements
-thought 4: Explore possible approaches
-thought 5: Comparative analysis of each approach
-thought 6: Select 3 options and detailed pros/cons
-thought 7: Recommendation and rationale
-```
-
-### Key principles
-
-```text
-✅ Output thinking process to actually think
-✅ Estimate high if complexity uncertain (can expand 5→7)
-✅ Concrete analysis needed per thought (no abstract descriptions)
-✅ Revise previous thoughts with isRevision if needed
-```
-
-</thinking_strategy>
-
----
-
-<codebase_exploration>
-
-## Task Subagent Usage
-
-### Subagent selection
-
-| subagent_type | Purpose | Example |
-|---------------|---------|---------|
-| **Explore** | Understand codebase structure, explore related files | "Analyze current auth structure" |
-| **Plan** | Establish implementation strategy, step-by-step planning | "Create migration plan" |
-| **general-purpose** | Complex analysis, multi-system correlation | "Analyze dependencies between modules" |
-
-### Task call patterns
-
-**Single exploration:**
-
-```typescript
-Task({
-  subagent_type: 'Explore',
-  description: 'Analyze current auth structure',
-  prompt: `
-    Understand current authentication-related code structure.
-    - Libraries in use
-    - Session management approach
-    - Files needing modification
-  `
-})
-```
-
-**Parallel exploration (complex cases):**
-
-```typescript
-// Multiple Task calls in single message
-Task({
-  subagent_type: 'Explore',
-  prompt: 'Analyze frontend auth structure'
-})
-
-Task({
-  subagent_type: 'Explore',
-  prompt: 'Analyze backend API auth endpoints'
-})
-
-Task({
-  subagent_type: 'Explore',
-  prompt: 'Analyze database session schema'
-})
-
-// → Combine results then organize options
-```
-
-### Exploration checklist
-
-```text
-✅ Understand current implementation approach
-✅ Libraries/framework versions in use
-✅ Related file and directory locations
-✅ Dependencies and related modules
-✅ Existing constraints (security, performance, compatibility)
-```
-
-</codebase_exploration>
-
----
-
-<option_presentation>
-
-## Option presentation format
-
-### Present 3 options (standard)
-
-```markdown
-## Analysis Results
-
-### Option 1: [Option name] (Recommended)
-
-**Approach:**
-- Description 1
-- Description 2
-
-| Pros | Cons |
-|------|------|
-| Pro 1 | Con 1 |
-| Pro 2 | Con 2 |
-
-**Impact scope:**
-- Files: `src/auth/`, `src/api/`
-- Expected change size: Medium
-- Risk: Low
-
----
-
-### Option 2: [Option name]
-
-**Approach:**
-...
-
-| Pros | Cons |
-|------|------|
-| ... | ... |
-
-**Impact scope:**
-...
-
----
-
-### Option 3: [Option name]
-
-**Approach:**
-...
-
----
-
-## Recommendation and rationale
-
-I recommend Option 1.
-- Rationale 1
-- Rationale 2
-
-Which option would you choose? (1/2/3)
-```
-
-### Present 2 options (when choices are clear)
-
-```markdown
-## Analysis Results
-
-There are two approaches:
-
-### Option A: [Option name]
-...
-
-### Option B: [Option name]
-...
-
-Which option would you choose? (A/B)
-```
-
-</option_presentation>
-
----
-
-<document_generation>
-
-## Document generation
-
-### Automatic document creation
-
-After user selects an option, automatically create a plan document at `.claude/plans/[feature-name].md`.
-
-```
-You selected Option [N].
-
-Creating plan document at .claude/plans/[feature-name].md...
-```
-
-### Plan document template
-
-**File location:** `.claude/plans/[feature-name].md`
-
-```markdown
-# [Feature name] Implementation Plan
-
-## Overview
-
-**Goal:** [What will be achieved]
-**Selected approach:** [Option N]
-**Expected impact scope:** [Files/modules list]
-
-## Current state
-
-- Current structure description
-- Related code location
-- Existing constraints
-
-## Implementation stages
-
-### Stage 1: [Stage name]
-
-**Tasks:**
-- [ ] Task 1
-- [ ] Task 2
-
-**Changed files:**
-- `src/file1.ts`
-- `src/file2.ts`
-
-### Stage 2: [Stage name]
-
-**Tasks:**
-- [ ] Task 3
-
-**Changed files:**
-- `src/file3.ts`
-
-### Stage 3: [Stage name]
-...
-
-## Considerations
-
-### Risks
-
-| Risk | Mitigation |
-|------|-----------|
-| Risk 1 | Plan 1 |
-| Risk 2 | Plan 2 |
-
-### Dependencies
-
-- External libraries: [List]
-- Other systems: [List]
-
-### Rollback plan
-
-How to rollback if issues occur.
-
-## Validation methods
-
-- Test item 1
-- Test item 2
-- Integration test
-
-## References
-
-- Related document links
-- Reference materials
-```
-
-</document_generation>
-
----
-
-<validation>
-
-## Validation checklist
-
-Before execution:
-
-```text
-✅ Verify ARGUMENT (ask if missing)
-✅ Sequential Thinking minimum 3 steps
-✅ Explore codebase with Task (Explore)
-✅ Minimum 2 options, recommended 3
-✅ List pros/cons for each option
-✅ Present impact scope and estimated work
-```
-
-Absolutely forbidden:
-
-```text
-❌ Start analysis without ARGUMENT
-❌ Use Edit tool (code modification forbidden)
-❌ Sequential Thinking less than 3 steps
-❌ Present only 1 option
-❌ Suggest options by guessing without code exploration
-❌ Start implementation without user choice
-❌ List options without pros/cons
-```
-
-</validation>
-
----
-
-<examples>
-
-## Real-world examples
-
-### Example 1: Change auth system
-
-```bash
-User: /plan Change user authentication from JWT to session-based
-
-1. Sequential Thinking (7 steps):
-   thought 1: "Auth system change - high complexity, multiple module impact"
-   thought 2: "Current JWT implementation analysis: src/auth/jwt.ts, token validation middleware"
-   thought 3: "Constraints: maintain existing user sessions, Redis/DB choice"
-   thought 4: "Approaches: incremental migration, full replacement, hybrid"
-   thought 5: "Incremental has low risk, hybrid high complexity"
-   thought 6: "Select 3 options and analyze pros/cons"
-   thought 7: "Recommend incremental migration - easy rollback, step-by-step validation"
-
-2. Task exploration:
-   Task (Explore): "Analyze current JWT auth implementation"
-   → Understand src/auth/, src/middleware/, API endpoints
-
-3. Present options:
-   Option 1: Incremental migration (Recommended)
-   - Pro: Easy rollback, low risk
-   - Con: Longer implementation time
-
-   Option 2: Full replacement
-   - Pro: Clean structure
-   - Con: High risk
-
-   Option 3: Hybrid approach
-   - Pro: Flexibility
-   - Con: Increased complexity
-
-4. User choice: 1
-
-5. Automatically create .claude/plans/session-auth.md
-```
-
-### Example 2: Add real-time notifications
-
-```bash
-User: /plan Add real-time notification feature
-
-1. Sequential Thinking (5 steps):
-   thought 1: "Real-time notifications - medium complexity, new feature"
-   thought 2: "Current communication: REST API, no polling"
-   thought 3: "Approaches: WebSocket, SSE, Long Polling, Firebase"
-   thought 4: "WebSocket bidirectional, SSE unidirectional but simpler"
-   thought 5: "Recommend WebSocket, polling inefficient"
-
-2. Task exploration:
-   Task (Explore): "Analyze current API structure and client communication"
-
-3. Options:
-   Option 1: WebSocket (Recommended)
-   Option 2: Server-Sent Events
-   Option 3: Short Polling
-
-4. After user selection, automatically create plan document
-```
-
-### Example 3: Simple refactoring
-
-```bash
-User: /plan Convert utils functions to TypeScript
-
-1. Sequential Thinking (3 steps):
-   thought 1: "Simple refactoring - simple, 1-2 files"
-   thought 2: "Need to analyze current utils.js"
-   thought 3: "Type definition → conversion → test validation"
-
-2. Task exploration:
-   Read: src/utils.js
-   Grep: search utils usages
-
-3. Options:
-   Option A: Incremental conversion (per file)
-   Option B: Batch conversion
-
-4. After user choice → create plan document → implement
-```
-
-</examples>