Claude Code Configuration
December 14, 2025 · View on GitHub
Personal Claude Code configuration repository containing agents, skills, commands, and coding standards for AI-assisted development workflows.
Overview
This repository provides a comprehensive configuration for Claude Code, Anthropic's CLI for Claude. It includes:
- 100+ specialized agents organized by domain
- 60+ reusable skills for common development tasks
- Custom slash commands for streamlined workflows
- Coding standards for multiple languages and frameworks
Repository Structure
.claude/
├── CLAUDE.md # Global instructions & behavior rules
├── agents/ # Specialized AI agents by category
│ ├── 01-core-development/
│ ├── 02-language-specialists/
│ ├── 03-infrastructure/
│ ├── 04-quality-security/
│ ├── 05-data-ai/
│ ├── 06-developer-experience/
│ ├── 07-specialized-domains/
│ ├── 08-business-product/
│ ├── 09-meta-orchestration/
│ └── 10-research-analysis/
├── skills/ # Reusable skill definitions
├── commands/ # Custom slash commands
│ └── git/ # Git workflow commands
├── includes/ # Language & framework standards
└── docs/ # Additional documentation
Agents
Agents are specialized AI assistants with domain expertise. Each agent has:
- Defined tools and capabilities
- Model preferences
- Contextual expertise
Categories
| Category | Description | Examples |
|---|---|---|
| Core Development | Full-stack & platform development | backend-developer, frontend-developer, fullstack-developer |
| Language Specialists | Language-specific expertise | python-pro, typescript-pro, golang-pro, rust-engineer |
| Infrastructure | DevOps, cloud, & platform | kubernetes-specialist, terraform-engineer, cloud-architect |
| Quality & Security | Testing, security, code review | security-engineer, qa-expert, penetration-tester |
| Data & AI | ML, data engineering, analytics | ml-engineer, data-scientist, llm-architect |
| Developer Experience | Tooling & workflows | cli-developer, build-engineer, documentation-engineer |
| Specialized Domains | Industry-specific | fintech-engineer, blockchain-developer, game-developer |
| Business & Product | Non-technical roles | product-manager, technical-writer, ux-researcher |
| Meta & Orchestration | Multi-agent coordination | workflow-orchestrator, task-distributor |
| Research & Analysis | Information gathering | research-analyst, competitive-analyst, market-researcher |
Skills
Skills are reusable capabilities that can be invoked during conversations. Notable skills include:
- Document Processing:
pdf,docx,xlsx,pptx - Media:
ai-multimodal,media-processing,image-enhancer - Development:
frontend-development,backend-development,databases - DevOps:
devops,chrome-devtools,webapp-testing - AI/ML:
anthropic-prompt-engineer,openai-prompt-engineer - Utilities:
changelog-generator,file-organizer,docs-seeker
Commands
Custom slash commands for common workflows:
Built-in Commands
| Command | Description |
|---|---|
/git:cm | Stage all files and create a commit |
/git:cp | Stage, commit, and push all changes |
/git:pr | Create a pull request |
/cr | Comprehensive code review with parallel specialist agents |
/cr-fx | Interactive remediation of code review findings |
/explore | Exhaustive codebase exploration (Opus 4.5) |
/deep-research | Multi-phase research protocol (Opus 4.5) |
Plugin: zircote/claude-spec
Project specification and implementation lifecycle management:
| Command | Description |
|---|---|
/cs:p | Strategic project planner with Socratic requirements elicitation |
/cs:i | Implementation progress tracker with PROGRESS.md checkpoints |
/cs:s | Project status and portfolio manager |
/cs:c | Project close-out and archival |
/cs:log | Toggle prompt capture logging |
/cs:wt:create | Create git worktree with Claude agent |
Install via /plugin command or add marketplace to settings:
{ "pluginMarketplaces": ["github:zircote/claude-spec"] }
Includes
Environment-specific coding standards automatically loaded based on context:
| Environment | File | Purpose |
|---|---|---|
| Python | includes/python.md | Python best practices & style |
| Go | includes/golang.md | Go idioms & patterns |
| React/TypeScript | includes/react.md | React & TS conventions |
| Git | includes/git.md | Version control workflows |
| Testing | includes/testing.md | Testing strategies |
| Documentation | includes/documentation.md | Doc standards |
| MCP | includes/mcp-reference.md | Model Context Protocol |
| Opus 4.5 | includes/opus-4-5.md | Opus 4.5 optimizations |
Configuration
Global Instructions (CLAUDE.md)
The CLAUDE.md file contains:
- Planning confidence requirements
- MCP query pagination patterns
- Memory agent usage guidelines
- API response validation rules
- Environment definitions
Ignored Files
The following are intentionally not tracked (see .gitignore):
- Session data (
history.jsonl,todos/,plans/) - Local settings (
settings.local.json,.env) - IDE configurations
- Plugins (reinstallable from marketplace)
- Project-specific data
Usage
- Clone to your home directory as
~/.claude/ - Claude Code automatically loads
CLAUDE.mdfor global instructions - Agents and skills are available via the Task tool
- Slash commands work with
/command:namesyntax
Requirements
- Claude Code CLI
- Anthropic API access
- Optional: MCP servers for extended capabilities
License
Personal configuration - adapt for your own use.