- Implement resolveActorFromMetadata() to extract attacker from message metadata UUID - Implement resolveItemFromMetadata() to extract weapon/ability from attacker's inventory - Change label format from "Damage (Actor, Weapon)" to "Actor -> Weapon" (e.g., "Goblin -> Scimitar Slash") - Add support for identifiedInfo, metadata, flavor text, and speaker.alias in priority order - Add support for Critical and Nonlethal damage modifiers - Update version to 0.1.19 This properly resolves WHO did the damage (attacker name) and WHAT was used (weapon/ability name), addressing the missing damage details. Format matches reference macros (macro_activate-hp-tracking.js). 🤖 Generated with Claude Code Co-Authored-By: Claude <noreply@anthropic.com>
Claude Code Setup - Complete Claude Code Configuration
Production-ready Claude Code setup with comprehensive documentation Version: 3.0.0 | Last Updated: 2025-10-20
🚀 Quick Start
New to this project? Start here: QUICKSTART.md (5-minute read)
Need complete details? See: CLAUDE_CODE_SETUP_COMPLETE.md (comprehensive)
📋 What's Included
This project demonstrates a fully configured Claude Code setup with:
Core Features (100% Configured)
✅ 8 MCP Servers
- Serena (code navigation + persistent memory)
- Sequential Thinking (complex reasoning)
- Context7 (real-time library docs)
- Memory (knowledge graph)
- Playwright (browser automation)
- Windows MCP (desktop automation)
- Fetch (web scraping)
- Database Server
✅ 8 Specialized Agents
- Architect, Code Reviewer, Debugger
- Documentation Writer, Project Manager
- Refactoring Specialist, Security Analyst, Test Engineer
✅ 9 Slash Commands
/analyze,/review,/implement,/test/optimize,/explain,/scaffold,/setup-info,/adr
✅ 6 Output Styles
- Concise, Professional, Verbose
- Explanatory, Learning, Security Reviewer
✅ 6 Event Hooks
- Session lifecycle (start/end)
- Bash command interception
- File write logging
- User prompt tracking
- Stop tracking for summaries
✅ Complete Templates
- Agent Template, Command Template
- Skill Template, Output Style Template
- CLAUDE.md Project Template
✅ Automatic Status Summaries
- Every task completion includes detailed summary
- Shows agents, commands, MCP servers used
- Lists all files modified
- Promotes transparency and learning
📚 Documentation
For Users
| Document | Purpose | Read Time |
|---|---|---|
| QUICKSTART.md | Get started quickly | 5 min |
| CLAUDE_CODE_SETUP_COMPLETE.md | Complete documentation | 30 min |
| MCP_SERVERS_GUIDE.md | Complete MCP server documentation | 20 min |
| CLAUDE.md | Project instructions for Claude | Reference |
For Developers
| Document | Purpose |
|---|---|
| CLAUDE_TEMPLATE.md | Template for new projects |
| .claude/TEMPLATES_README.md | Master template guide |
| .claude/agents/MCP_USAGE_TEMPLATES.md | MCP usage templates for agents |
| .claude/TEMPLATE_CAPABILITIES_ANALYSIS.md | Template review & missing features |
Specialized Guides
| Document | Purpose |
|---|---|
| MCP_SERVERS_GUIDE.md | Complete MCP server documentation |
| .claude/CHECKPOINTING_GUIDE.md | Session checkpointing |
| .claude/PLUGIN_SETUP.md | Plugin marketplace |
| .claude/STATUS_LINE_SETUP.md | Status line customization |
🎯 Key Capabilities
Intelligent Code Navigation (Serena MCP)
# Find symbols
find_symbol("UserService")
# Find references
find_referencing_symbols("authenticate", "src/auth/")
# Store persistent knowledge
write_memory("adr-001-architecture", "Decision: Microservices...")
Specialized Agents (8 Available)
# Automatic invocation
> "I need to design a microservices architecture"
# → Architect agent activated
# Manual invocation
> "Use the security-analyst agent to review this code"
Development Workflows (8 Commands)
/analyze src/ # Code analysis
/review src/ # Code review
/implement feature # Build features
/test # Run tests
/optimize file.ts # Performance tuning
Extended Thinking
> "Think hard about the best database architecture"
> "Ultrathink: How to optimize this algorithm?"
Custom Output Modes (6 Styles)
/output-style learning # Interactive learning
/output-style security-reviewer # Security focus
/output-style explanatory # Educational insights
📁 Project Structure
Claude Code Setup/
├── README.md # This file
├── QUICKSTART.md # Quick start guide (5 min)
├── CLAUDE_CODE_SETUP_COMPLETE.md # Complete documentation (30 min)
├── CLAUDE.md # Project instructions
├── CLAUDE_TEMPLATE.md # Project template
│
├── .mcp.json # MCP servers configuration
│
└── .claude/ # Main configuration directory
├── settings.json # Shared configuration
├── settings.local.json # Personal configuration
│
├── agents/ # 8 specialized agents
│ ├── architect.md
│ ├── code-reviewer.md
│ ├── debugger.md
│ ├── documentation-writer.md
│ ├── project-manager.md
│ ├── refactoring-specialist.md
│ ├── security-analyst.md
│ ├── test-engineer.md
│ └── .AGENT_TEMPLATE.md
│
├── commands/ # 9 slash commands
│ ├── adr.md
│ ├── analyze.md
│ ├── explain.md
│ ├── implement.md
│ ├── optimize.md
│ ├── review.md
│ ├── scaffold.md
│ ├── setup-info.md
│ ├── test.md
│ └── .COMMANDS_TEMPLATE.md
│
├── output-styles/ # 6 custom styles
│ ├── concise.md
│ ├── professional.md
│ ├── verbose.md
│ ├── explanatory.md
│ ├── learning.md
│ ├── security-reviewer.md
│ └── .OUTPUT_STYLES_TEMPLATE.md
│
├── skills/ # Skills directory
│ └── .SKILL_TEMPLATE.md
│
├── hooks/ # 5 event hooks
│ ├── session-start.sh
│ ├── session-end.sh
│ ├── pre-bash.sh
│ ├── post-write.sh
│ └── user-prompt-submit.sh
│
├── tools/ # Utility scripts
│ ├── start-memory.ps1
│ └── statusline.sh
│
├── logs/ # Session logs
│
└── [Documentation Files]
├── TEMPLATES_README.md
├── MCP_CORRECT_USAGE_GUIDE.md
├── TEMPLATE_CAPABILITIES_ANALYSIS.md
├── CHECKPOINTING_GUIDE.md
├── PLUGIN_SETUP.md
└── STATUS_LINE_SETUP.md
🚦 Getting Started
1. First Time Setup (1 minute)
# Clone or open this project
cd "Claude Code Setup"
# Start Claude Code
claude
# Verify setup
> /setup-info
2. Try Basic Commands (2 minutes)
# Get help
> /help
# See what's configured
> /setup-info
# Try a command
> /analyze src/
# Try an agent
> "Use the architect agent to explain the project structure"
3. Learn Core Features (5 minutes)
Read QUICKSTART.md for:
- Essential commands
- MCP server usage
- Agent invocation
- Output styles
- Memory system
4. Deep Dive (30 minutes)
Read CLAUDE_CODE_SETUP_COMPLETE.md for complete details on:
- All MCP servers and capabilities
- Agent configuration and usage
- Slash command reference
- Hooks system
- Templates
- Advanced features
- Best practices
💡 Common Use Cases
Feature Development
# 1. Architecture
> "Use architect agent to design payment integration"
# 2. Implement
> /implement Stripe payment integration
# 3. Test
> /test src/payments/
# 4. Review
> /review src/payments/
# 5. Document
> "Use documentation-writer agent to document payment flow"
Bug Fixing
# 1. Debug
> "Use debugger agent: [paste error]"
# 2. Extended thinking (for complex bugs)
> "Think hard about this race condition"
# 3. Review fix
> /review [fixed file]
Code Review
# 1. Standard review
> /review src/
# 2. Security check
> "Use security-analyst agent to check vulnerabilities"
# 3. Refactoring suggestions
> "Use refactoring-specialist agent for improvements"
Learning Codebase
# 1. Use explanatory style
> /output-style explanatory
# 2. High-level questions
> "Explain the architecture of this project"
# 3. Deep dive with Serena
> get_symbols_overview("src/core/engine.ts")
# 4. Store learnings
> write_memory("architecture-overview", "The system uses...")
🎓 Learning Path
Beginner (Day 1)
- Read QUICKSTART.md
- Try basic commands (
/help,/setup-info) - Experiment with one agent
- Try one output style
Intermediate (Week 1)
- Use all slash commands
- Work with multiple agents
- Try extended thinking
- Use Serena MCP for code navigation
Advanced (Month 1)
- Create custom commands
- Build custom agents
- Configure hooks for your workflow
- Leverage all MCP servers
🔧 Customization
Add a Custom Command
# Copy template
cp .claude/commands/.COMMANDS_TEMPLATE.md .claude/commands/my-command.md
# Edit and configure
# Use it
> /my-command
Add a Custom Agent
# Copy template
cp .claude/agents/.AGENT_TEMPLATE.md .claude/agents/my-agent.md
# Configure and use
> "Use the my-agent agent to..."
Add a Custom Output Style
# Copy template
cp .claude/output-styles/.OUTPUT_STYLES_TEMPLATE.md .claude/output-styles/my-style.md
# Activate
> /output-style my-style
See .claude/TEMPLATES_README.md for detailed guides.
🤝 Team Setup
Sharing This Setup
# Everything is in git, just commit and push
git add .claude/ .mcp.json CLAUDE.md
git commit -m "Add Claude Code configuration"
git push
# Team members pull and get full setup
git pull
Team Customization
Shared (in git):
.claude/settings.json.claude/agents/.claude/commands/.mcp.jsonCLAUDE.md
Personal (not in git):
.claude/settings.local.json~/.claude/(user-wide)
📊 Setup Completeness
Current: ~95% of Claude Code capabilities implemented
Implemented:
- ✅ MCP servers
- ✅ Agents/Subagents
- ✅ Slash commands
- ✅ Output styles
- ✅ Hooks
- ✅ Templates
- ✅ Memory system
- ✅ Extended thinking
- ✅ Plan mode
- ✅ Checkpointing
- ✅ Plugin marketplace
Optional (not implemented):
- ⚠️ Enterprise SSO/Analytics
- ⚠️ Headless mode (CI/CD)
- ⚠️ Network proxies/certificates
🖥️ Platform Notes
Windows-Only Components
The following components are Windows-specific and will not work on Mac/Linux:
-
Windows MCP (.windows-mcp/) - Desktop automation using Python/uv
- Provides Windows UI automation, PowerShell execution, clipboard operations
- Impact on other platforms: Will show error on startup but won't affect other MCP servers
- Solution for Mac/Linux: Ignore Windows MCP errors or disable in
.mcp.json
-
PowerShell Scripts:
- .claude/tools/start-memory.ps1 - Memory MCP launcher
- Various hook scripts using PowerShell commands
- Alternative: Create bash equivalents for cross-platform support
Cross-Platform Components
All other MCP servers work on all platforms:
- ✅ Serena (Python/uvx)
- ✅ Sequential Thinking (Node.js/npx)
- ✅ Database Server (Node.js/npx)
- ✅ Context7 (Node.js/npx)
- ✅ Memory (works via Node.js on all platforms - PowerShell launcher is Windows-only convenience)
- ✅ Fetch (Python/uvx)
- ✅ Playwright (Node.js/npx)
Disabling Windows MCP on Mac/Linux
If you're on Mac/Linux and want to disable Windows MCP to avoid startup errors:
Option 1: Comment out in .mcp.json
{
"mcpServers": {
// "windows-mcp": {
// "command": "uv",
// "args": ["--directory", "./.windows-mcp", "run", "main.py"]
// },
// ... other servers
}
}
Option 2: Disable in settings
// In .claude/settings.json or .claude/settings.local.json
{
"mcpServers": {
"windows-mcp": {
"enabled": false
}
}
}
🐛 Troubleshooting
Quick fixes:
# Agent not working?
> "Use the [agent-name] agent to..." # Manual invocation
# Command not found?
> /help # List available commands
# MCP server failed?
cat .mcp.json | jq '.mcpServers' # Check configuration
# Permission denied?
cat .claude/settings.json | jq '.permissions' # Check permissions
# Windows MCP error on Mac/Linux?
# This is normal - see Platform Notes above
See CLAUDE_CODE_SETUP_COMPLETE.md for detailed troubleshooting.
📖 Resources
Official Documentation
- Claude Code Docs: https://docs.claude.com/en/docs/claude-code/
- GitHub Issues: https://github.com/anthropics/claude-code/issues
This Project
- Quick Start: QUICKSTART.md
- Complete Guide: CLAUDE_CODE_SETUP_COMPLETE.md
- MCP Servers: MCP_SERVERS_GUIDE.md
- Templates: .claude/TEMPLATES_README.md
- MCP Templates: .claude/agents/MCP_USAGE_TEMPLATES.md
🎉 Success Checklist
After setup, you should be able to:
- Run
/setup-infoand see full configuration - Use slash commands like
/analyze,/review,/test - Invoke agents automatically or manually
- Change output styles with
/output-style - Use Serena MCP for code navigation
- Store persistent memories
- Use extended thinking for complex problems
- Access checkpoints with ESC ESC
Ready? Start with QUICKSTART.md!
📝 Version History
- v3.0.0 (2025-10-20): Complete documentation overhaul, consolidated docs, added /adr command
- v2.0.0 (2025-10-17): Added output styles, plugins, status line
- v1.0.0: Initial comprehensive setup
📄 License
This configuration setup is provided as-is for educational and development purposes. Feel free to use, modify, and share.
Questions? Check the documentation files or see troubleshooting
Ready to start? → QUICKSTART.md