ccstart 2.0.0

package/README.md

@@ -0,0 +1,346 @@
+# ccstart
+
+[![npm version](https://img.shields.io/npm/v/ccstart.svg)](https://www.npmjs.com/package/ccstart)
+[![PRs Welcome](https://img.shields.io/badge/PRs-welcome-brightgreen.svg)](https://github.com/vichannnnn/claude-code/pulls)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+[![npm downloads](https://img.shields.io/npm/dm/ccstart.svg)](https://www.npmjs.com/package/ccstart)
+
+Quick setup for Claude Code projects with built-in agents, ticket system, planning tools and agent orchestration workflow.
+
+## Installation
+
+```bash
+# Create a new project
+npx ccstart my-project
+
+# Setup in current directory
+npx ccstart .
+
+# Install globally (optional)
+npm install -g ccstart
+ccstart my-project
+```
+
+## Command Line Options
+
+```bash
+ccstart [project-name] [options]
+
+Options:
+  --force, -f     Skip all prompts and overwrite existing files
+  --dry-run, -d   Show what would be done without making changes
+  --agents        Interactive agent selection mode
+  --all-agents    Include all agents without prompting
+  --no-agents     Skip agent selection entirely
+  --help, -h      Show help message
+
+Examples:
+  ccstart                    # Create in current directory
+  ccstart my-project         # Create in new directory
+  ccstart . --force          # Overwrite files in current directory
+  ccstart my-app --dry-run   # Preview changes without creating files
+  ccstart --agents           # Interactive agent selection only
+  ccstart my-app --all-agents # Include all agents automatically
+```
+
+## What's Included
+
+The boilerplate template creates:
+- **CLAUDE.md** - Project instructions for Claude (references claude/ subdirectories)
+- **claude/** - All Claude-related files in one organized directory:
+  - **claude/agents/** - Specialized AI agents (planner, coder, checker, etc.)
+  - **claude/tickets/** - Task tracking system
+  - **claude/plans/** - Project planning documents
+  - **claude/docs/** - Documentation with ROADMAP.md and agent-orchestration.md
+- **.claude/** - Claude Code specific directory structure (created automatically)
+
+## Key Features
+
+- 🎯 **Interactive Agent Selection** - Choose which agents to include during setup
+- 🔄 **Agent Orchestration Workflows** - Pre-defined workflows that automatically coordinate multiple agents for complex tasks
+- 🔒 **Smart Conflict Resolution** - Handle existing files with skip/rename/overwrite options
+- 📁 **.claude Directory Support** - Automatic creation of Claude Code directory structure
+- 🏃 **Dry Run Mode** - Preview changes before applying them
+- ⚡ **Force Mode** - Skip all prompts for automated workflows
+- 🎨 **Standalone Agent Preview** - Explore available agents without setup
+- 🎭 **Systematic Task Execution** - Agents work in sequence, building on each other's outputs
+
+## Project Structure
+
+- **templates/** - The boilerplate templates that will be copied to user's project
+- **bin/** - CLI executable for ccstart
+- **agents/**, **docs/**, **plans/**, **tickets/** - Working files for ccstart development (not part of the template)
+
+## Usage
+
+1. Create a new project:
+   ```bash
+   npx ccstart my-awesome-project
+   cd my-awesome-project
+   ```
+
+2. Edit `CLAUDE.md` with your project-specific instructions
+
+3. Update `claude/docs/ROADMAP.md` with your project goals
+
+4. Start creating tickets in `claude/tickets/` directory
+
+## Claude Code Integration
+
+ccstart seamlessly integrates with Claude Code:
+
+- **Automatic Detection** - Checks if Claude Code is initialized in your project
+- **.claude Directory** - Creates Claude Code directory structure automatically
+- **Agent Installation** - Copies agents to `.claude/agents/` for native Claude Code support
+- **Smart Initialization** - Offers to create `.claude` directory if Claude Code isn't detected
+
+When running in the current directory (`npx ccstart .`), it will check for Claude Code and provide instructions if not found. For new projects, ccstart will remind you to initialize Claude Code after setup.
+
+## Using Agents
+
+The boilerplate includes several specialized agents in the `agents/` directory:
+
+- **planner.md** - Strategic planning and task breakdown
+- **coder.md** - Implementation and development
+- **checker.md** - Testing and quality assurance
+- **researcher.md** - Research and information gathering
+- **blockchain.md** - Web3 and smart contract development
+- **frontend.md** - UI/UX and frontend development
+- **backend.md** - API design and server-side development
+- **shadcn.md** - shadcn/ui component development
+
+### Interactive Agent Selection
+
+During setup, ccstart will prompt you to select which agents to include:
+
+```bash
+🤖 Select agents to include in your Claude Code project
+
+Use arrow keys to navigate, space to select/deselect, 'a' to toggle all
+
+Choose your agents:
+◯ backend - Backend development specialist for API design and server optimization
+◯ blockchain - Web3 and smart contract development expert
+◯ checker - Quality assurance and code review specialist
+◯ coder - Expert software developer for implementation
+◯ frontend - Frontend development specialist for UI/UX
+◯ planner - Strategic planning specialist for complex problems
+◯ researcher - Research specialist for documentation and code analysis
+◯ shadcn - shadcn/ui component library expert
+```
+
+You can also use flags to control agent selection:
+- `--all-agents` - Include all available agents
+- `--no-agents` - Skip agent selection entirely
+- `--agents` - Preview available agents without creating a project
+
+Example:
+```bash
+# Include all agents without prompting
+npx ccstart my-project --all-agents
+
+# Create project without any agents
+npx ccstart my-project --no-agents
+
+# Preview available agents only
+npx ccstart --agents
+```
+
+### Conflict Resolution
+
+When existing files are detected, ccstart offers smart conflict resolution:
+
+```bash
+⚠️  File conflicts detected. You will be asked how to handle each category.
+
+📄 CLAUDE.md conflicts:
+  - CLAUDE.md
+
+Conflict resolution options:
+  1) skip      (s) - Keep your existing files
+  2) rename    (r) - Save template files with -ccstart suffix
+  3) overwrite (o) - Replace with template versions
+
+Your choice for CLAUDE.md [s/r/o]: 
+```
+
+Categories are handled separately:
+- **CLAUDE.md** - Project instructions (warns before overwriting)
+- **Agents** - AI agent files
+- **Documentation** - docs/ folder files
+- **Plans** - plans/ folder files
+- **Tickets** - tickets/ folder files
+
+## Agent Orchestration Showcase
+
+The template includes powerful agent orchestration workflows that guide Claude through complex tasks systematically. Here are real-world examples:
+
+### 🚀 Feature Development Example
+
+**Scenario**: Building a user authentication system
+
+```bash
+# In Claude Code, you would say:
+"I need to add user authentication to my app. Use the feature development workflow to guide me through this."
+
+# Claude will automatically orchestrate:
+1. Researcher Agent → Analyzes existing codebase, identifies auth patterns
+2. Planner Agent → Creates detailed implementation plan with JWT, sessions, etc.
+3. Coder Agent → Implements auth endpoints, middleware, and UI components
+4. Checker Agent → Tests security, validates implementation, checks edge cases
+```
+
+### 🐛 Bug Fix Example
+
+**Scenario**: Users report login failures after 5 minutes
+
+```bash
+# You tell Claude:
+"Users are getting logged out after 5 minutes. Use the bug fix workflow to investigate and fix this."
+
+# Claude orchestrates:
+1. Researcher Agent → Investigates session handling, token expiry, logs
+2. Coder Agent → Fixes token refresh logic, updates session timeout
+3. Checker Agent → Verifies fix, tests edge cases, ensures no regressions
+```
+
+### 🏗️ API Development Example
+
+**Scenario**: Building a REST API for a blog platform
+
+```bash
+# You request:
+"Help me build a complete REST API for blog posts with CRUD operations. Follow the API development workflow."
+
+# Claude coordinates:
+1. Planner Agent → Designs RESTful endpoints, database schema, auth strategy
+2. Backend Agent → Implements controllers, models, middleware, validation
+3. Frontend Agent → Creates API client, integration examples (if needed)
+4. Checker Agent → Tests all endpoints, validates OpenAPI spec, security audit
+```
+
+### 🎨 UI Component Example
+
+**Scenario**: Creating a responsive dashboard with shadcn/ui
+
+```bash
+# You ask:
+"Create a dashboard with charts and stats cards using shadcn. Use the UI component workflow."
+
+# Claude executes:
+1. Frontend Agent → Designs component structure, responsive layout
+2. Shadcn Agent → Implements with shadcn/ui components, themes, animations
+3. Checker Agent → Tests responsiveness, accessibility, performance
+```
+
+### 🔍 Quality Assurance Example
+
+**Scenario**: Comprehensive review before deployment
+
+```bash
+# You request:
+"Run full QA on the authentication feature we just built."
+
+# Claude performs:
+1. Researcher Agent → Analyzes all changes, identifies test requirements
+2. Checker Agent → Runs security scans, performance tests, accessibility checks
+3. Coder Agent → Fixes any issues found
+4. Checker Agent → Final validation, generates QA report
+```
+
+### 💡 How to Use Agent Orchestration
+
+1. **Read the orchestration guide**:
+   ```
+   "Read docs/agent-orchestration.md to understand available workflows"
+   ```
+
+2. **Choose your workflow**:
+   - Feature Development: Complex new features
+   - Bug Fix: Investigating and fixing issues
+   - Refactoring: Improving code quality
+   - API Development: Building APIs
+   - UI Components: Frontend development
+   - Blockchain: Web3 features
+   - QA: Quality assurance
+
+3. **Let Claude guide you**:
+   ```
+   "I need to [your task]. Which workflow should we use?"
+   ```
+
+4. **Follow the flow**:
+   - Claude will automatically use agents in the correct sequence
+   - Each agent builds on the previous one's work
+   - You get systematic, thorough results
+
+The orchestration ensures nothing is missed and follows best practices automatically!
+
+## Getting Started with Claude Code
+
+After setting up your project with `ccstart`:
+
+1. **Open Claude Code** in your project directory:
+   ```bash
+   cd my-awesome-project
+   claude
+   ```
+
+2. **Let Claude understand your project** by asking:
+   - "Read the CLAUDE.md file to understand this project"
+   - "Check the roadmap in docs/ROADMAP.md"
+   - "What agents are available in the agents directory?"
+   - "Read the README files in docs, tickets, and plans folders to understand the workflow"
+
+3. **Start working** with Claude:
+   - Use the planner agent: "Use the planner agent to help me design a user authentication system"
+   - Create tickets: "Create a ticket for implementing user login"
+   - Implement features: "Use the coder agent to implement the login functionality"
+   - Review code: "Use the checker agent to review the code we just wrote"
+
+4. **Important setup steps**:
+   - **Update ROADMAP.md**: Define your project's goals, features, and development phases
+   - **Read folder documentation**: Each folder (docs/, tickets/, plans/) has a README explaining its purpose and format
+   - **Customize CLAUDE.md**: Add project-specific instructions, commands, and context
+
+5. **Workflow tips**:
+   - Always start with planning for complex features
+   - Create tickets to track progress
+   - Use the appropriate agent for each task
+   - Keep CLAUDE.md updated with project-specific instructions
+   - Follow the workflow defined in docs/ROADMAP.md
+
+## Features
+
+### Core Features
+- ✅ Pre-configured project structure for Claude Code
+- 🤖 Multiple specialized agents for different tasks
+- 🎫 Built-in ticket and planning system
+- 📋 Ready-to-use boilerplate with best practices
+- 📁 Automatic .claude directory integration
+
+### CLI Features
+- 🎯 Interactive agent selection with descriptions
+- 🔄 Smart conflict resolution (skip/rename/overwrite)
+- 👀 Dry-run mode to preview changes
+- ⚡ Force mode for CI/CD pipelines
+- 🎨 Standalone agent preview mode
+- 🛡️ Security validations for file operations
+
+### Agent Orchestration
+The template includes `docs/agent-orchestration.md` which defines workflows for:
+- **Feature Development** - Researcher → Planner → Coder → Checker
+- **Bug Fixes** - Researcher → Coder → Checker
+- **Refactoring** - Researcher → Planner → Coder → Checker
+- **API Development** - Planner → Backend → Frontend → Checker
+- **UI Components** - Frontend → Shadcn → Checker
+- **Blockchain** - Planner → Blockchain → Checker
+- **Quality Assurance** - Researcher → Checker → Coder → Checker
+
+## Credits
+
+Born from our discussions in TechOverflow with [vichannnnn](https://github.com/vichannnnn) and [nasdin](https://github.com/nasdin)
+
+## License
+
+MIT