v2.1.0: chore: Bump version to 2.1.0

🎯 New Features

Repository Selection During Init

• Checkbox UI for selecting which repositories to include
• Space to toggle, Enter to confirm
• Default: all repositories checked
• Validation: at least one repository required

? 포함할 레포지토리를 선택하세요:
  ◉ backend-api (백엔드)
  ◉ frontend-web (프론트엔드)
  ◯ test-project (제외)

✓ 2개 레포지토리 선택됨

Removed Inaccurate Tech Stack Inference

- Deleted 183 lines of inference code
- Init only detects repository type (frontend/backend/mobile/fullstack)
- AI analyzes tech stack accurately during setup phase

Before:
Stack: JavaScript, Express, Socket.IO  ← Often wrong

After:
AI가 분석할 내용:
  - 기술 스택
  - 프로젝트 설명
  - 주요 기능

💡 Why This Matters

Problem with v2.0.0:
- Early tech stack detection was inaccurate
- Misleading information during init
- Complex inference logic (200+ lines)

Solution in v2.1.0:
- No more guessing
- AI reads actual code → 100% accurate
- Simpler, faster init
- User chooses which repos to include

🔧 Enhanced README

- Step 5 highlighted with ⚠️ IMPORTANT
- Clear instructions with emojis (1️⃣ 2️⃣ 3️⃣)
- "Why this works" explanation added

📋 Full Changelog

Features:
- 7582245 feat: Add repository selection during init
- ded07f2 docs: Highlight Step 5 in README

Fixes:
- db63487 fix: Remove inferred tech stack from repository selection
- f70ab9c refactor: Remove tech stack inference completely

Chore:
- 573d4e4 chore: Bump version to 2.1.0

v2.0.1: refactor: Remove tech stack inference completely

v2.0.1 릴리즈 노트 (업데이트):

v2.0.1 - Clean Init & Accurate AI Analysis

🎯 Major Improvements

Removed Inaccurate Tech Stack Inference

• Deleted 183 lines of inference code
• Init now only detects repository type
• AI analyzes everything accurately during setup

Why this is better:

• Early inference was often wrong
• AI reads actual code → 100% accurate
• Faster init process
• Cleaner codebase

Repository Selection UI

• Checkbox interface for selecting repos
• Only show repo name and type
• No misleading tech stack info

🔧 Technical Changes

Removed:

• detectTechStack() function (~100 lines)
• detectDescription() function (~80 lines)
• All dependency scanning during init

Improved:

• SETUP_GUIDE.md now says "To be analyzed by AI"
• Types made optional (description?, techStack?)
• Simpler, more maintainable code

v2.0.0: feat: Enhance comment-based context management system

🚀 Major Changes

AI-Driven Setup Architecture

• Framework provider pattern: CodeSyncer provides structure (WHERE/HOW), AI provides content (WHAT)
• Simplified init command: Only generates SETUP_GUIDE.md
• AI reads guide and generates all .claude/ files interactively
• Complete rewrite of init.ts (513 → 189 lines)

Comment-Based Context Management System ⭐

• All context now managed through comments instead of separate documentation
• Removed unnecessary files: QUALITY_CHECKLIST, CODING_STANDARDS, EXAMPLES
• Enhanced COMMENT_GUIDE.md with comprehensive real-world examples
• 10 comment tags system (5 basic + 5 extended)

💡 Philosophy Shift

Before (v1.x)

❌ Direct file generation → User couldn't customize
❌ Separate long documentation → AI can't remember
❌ Static templates → One size fits all

After (v2.0.0)

✅ AI-driven setup → Analyzes YOUR code
✅ Comment-based context → Permanent record in code
✅ Interactive process → AI asks questions, never assumes
✅ Lightweight system → Only 4 essential files

📋 Comment Tag System (10 Tags)

Basic (5):

• @codesyncer-inference - Inference + rationale
• @codesyncer-decision - [Date] + decision
• @codesyncer-todo - Needs confirmation
• @codesyncer-context - Business context
• @codesyncer-rule - Special rule

Extended (5):

• @codesyncer-why - Detailed explanation
• @codesyncer-tradeoff - Pros and cons
• @codesyncer-alternative - Other options considered
• @codesyncer-pattern - Pattern name (reusable)
• @codesyncer-reference - Reference link

📝 Real-World Comment Examples

COMMENT_GUIDE.md now includes production-ready examples for:

1. Quality standards (timeout, retry, logging)
2. Complex business logic (discount calculation order)
3. Performance optimization (cursor pagination, caching)
4. Security requirements (OWASP, authentication)
5. Error handling (Circuit Breaker, Retry, Fallback)
6. Test strategy (AAA pattern, coverage goals)

🔄 Migration from v1.x

If upgrading from v1.x:

1. Backup existing .claude/ files
2. Run codesyncer init in workspace root
3. AI will generate new files based on YOUR actual code
4. Review and merge with your customizations
5. Start using enhanced comment tags

🎓 Quick Start

npm install -g codesyncer@latest
codesyncer init

Made with ❤️ by the CodeSyncer community

v1.1.0

🎯 Major Feature: Repository Name Analysis

New Features

• ✅ Smart Project Type Detection: Analyzes repository name keywords

  • Socket servers (*-socket, *-server) correctly detected as backend
  • Admin dashboards (admin-*, *-dashboard) detected as frontend
  • API servers (*-api, *-server) detected as backend

• ✅ Enhanced Description Generation

  • Uses repo name to create meaningful descriptions

Bug Fixes

• Fixed issue where all Next.js projects were classified as "fullstack"
• Socket.IO projects now correctly identified as backend

v1.0.5

🐛 Critical Bug Fix

Fixed

• ✅ Version Display Issue: --version now correctly shows the actual package version
  • Previously hardcoded to 1.0.0
  • Now dynamically reads from package.json
  • Users will see correct version after updates

Previous Updates

• v1.0.4: Added update guide to README
• v1.0.3: Smart auto-detection for descriptions and tech stack
• v1.0.2: Added README and LICENSE to npm package
• v1.0.1: Fixed package-lock.json for CI/CD

Testing

After installing this version:

npm install -g codesyncer@latest
codesyncer --version
# Should now show: 1.0.5

v1.0.3

🎯 Major Improvements

Auto-Detection Enhancements

• ✅ Smart Description Generation: Automatically detects and fills project descriptions

  • Reads from package.json if available
  • Generates meaningful descriptions based on tech stack
  • Example: "Real-time backend server with WebSocket" for socket projects

• ✅ Enhanced Tech Stack Detection:

  • Socket.IO / WebSocket support
  • Database detection (Prisma, MongoDB, PostgreSQL, MySQL)
  • Better framework recognition

User Experience

• No more empty description fields during setup
• Pre-filled tech stack based on dependencies
• Users can still modify auto-detected values

Bug Fixes

• Fixed issue where description was always empty
• Improved accuracy of project type detection

v1.0.2

🐛 Bug Fix

• Added README.md to npm package (fixes missing documentation on npm page)
• Added README.ko.md for Korean documentation
• Added LICENSE file to package

📦 Package Contents

Now includes:

• ✅ README.md (English)
• ✅ README.ko.md (Korean)
• ✅ LICENSE
• ✅ Binary files
• ✅ Templates

v1.0.1

🧪 Test Automated npm Deployment

Testing the automated CI/CD pipeline:

• GitHub Actions → npm publish
• Automated deployment on release creation

Changes

• Added package-lock.json for CI/CD
• Fixed test workflow
• Cleaned up all references