Co-Pilot / 辅助式
更新于 a month ago

claude-settings

Eeddiebe147
0.0k
eddiebe147/claude-settings
76
Agent 评分

💡 摘要

一个用于在机器和团队之间同步Claude代码设置的配置库。

🎯 适合人群

软件开发人员DevOps工程师项目经理技术团队负责人AI研究人员

🤖 AI 吐槽:看起来很能打,但别让配置把人劝退。

安全分析中风险

风险:Medium。建议检查:是否执行 shell/命令行指令;是否发起外网请求(SSRF/数据外发);文件读写范围与路径穿越风险;依赖锁定与供应链风险。以最小权限运行,并在生产环境启用前审计代码与依赖。

Claude Code Settings - ID8Labs

Personal Claude Code configuration for cross-machine sync and team sharing.

Quick Start

New Machine Setup

cd ~ git clone https://github.com/eddiebe147/claude-settings.git .claude

Sync Between Machines

# Pull latest cd ~/.claude && git pull # Push changes cd ~/.claude && git add . && git commit -m "update" && git push

What's Inside

| Folder | Purpose | Count | |--------|---------|-------| | agents/ | Specialized AI agents for specific domains | 38+ | | skills/ | Reusable skill modules (slash commands) | 21 | | hooks/ | Automation scripts (pre/post actions) | 26 | | commands/ | Quick slash commands | 21 | | plugins/ | Claude Code plugin configs | - |

Core Files

| File | Purpose | |------|---------| | CLAUDE.md | Global instructions for ALL projects | | settings.json | Claude Code preferences | | hooks.json | Hook configurations | | mcpServers.json | MCP server connections |


Agents

Specialized AI agents in agents/ folder. Each handles a specific domain.

Development Agents

| Agent | Description | |-------|-------------| | nextjs-senior-dev.md | Next.js 14+ App Router expert | | backend-architect.md | Backend system architecture | | database-migration-specialist.md | Database schema migrations |

MCP Ecosystem Agents

| Agent | Description | |-------|-------------| | mcp-server-architect.md | Design MCP servers | | mcp-protocol-specialist.md | Protocol compliance | | mcp-integration-engineer.md | Multi-server orchestration | | mcp-deployment-orchestrator.md | Production deployments | | mcp-security-auditor.md | Security reviews | | mcp-testing-engineer.md | Testing & QA | | mcp-registry-navigator.md | Registry discovery |

Business & Strategy Agents

| Agent | Description | |-------|-------------| | market-intelligence-analyst.md | Market research & trends | | steve-jobs-advisor.md | Visionary product guidance | | strategic-think-tank.md | Major business decisions | | operations-manager.md | Project coordination | | relationship-builder.md | Client relationships |

Content & Marketing Agents

| Agent | Description | |-------|-------------| | social-media-manager.md | Social content creation | | x-viral-optimizer.md | X/Twitter optimization | | nana-image-generator.md | AI image generation | | notebooklm-producer.md | Audio/video from docs | | reality-tv-beat-writer.md | Documentary production |

TMNT Dev Team (Fun Names!)

| Agent | Role | |-------|------| | leonardo-strategist-agent.md | Strategy & planning | | donatello-architect-agent.md | Technical architecture | | michelangelo-builder-agent.md | Implementation | | raphael-guardian-agent.md | Security & protection | | splinter-mentor-agent.md | Mentorship | | krang-brain-agent.md | AI/ML expertise | | bebop-devops-agent.md | DevOps | | rocksteady-database-agent.md | Database | | casey-tester-agent.md | Testing |

Full Projects

| Project | Description | |---------|-------------| | reddit-research/ | Market research agent using Reddit's JSON API |


Skills

Reusable modules in skills/ that extend Claude's capabilities.

Development Skills

| Skill | Purpose | |-------|---------| | nextjs-project-manager | Next.js 14+ project guidance | | api-design | RESTful API design | | database-design | Schema & Supabase | | supabase-expert | Supabase integration | | testing-qa | Test automation | | performance-optimization | Speed & Core Web Vitals | | state-management | React state (Zustand, Context) | | ui-builder | React + Tailwind + Shadcn |

Infrastructure Skills

| Skill | Purpose | |-------|---------| | deployment-automation | Vercel & CI/CD | | file-uploads | File handling & cloud storage | | error-monitoring | Logging & Sentry | | real-time-features | WebSockets & Supabase Realtime | | mcp-builder | Build MCP servers |

Business Skills

| Skill | Purpose | |-------|---------| | llc-ops | LLC operations for ID8Labs | | analytics-tracking | User analytics & A/B testing | | email-notifications | Transactional emails |


Hooks

Automation scripts in hooks/ that run before/after actions.

Quality Gates

| Hook | Trigger | Purpose | |------|---------|---------| | code-review-gate.sh | Before PR | Code quality check | | quality-gate-check.sh | Before commit | Standards enforcement | | pre-commit.sh | Before commit | Pre-commit checks | | post-commit.sh | After commit | Post-commit actions |

Workflow Automation

| Hook | Purpose | |------|---------| | auto-commit.sh | Automatic commits | | auto-git-workflow.sh | Git flow automation | | auto-documentation.sh | Doc generation | | enforce-feature-branch.sh | Branch protection | | suggest-feature-branch.sh | Branch suggestions |

Verification

| Hook | Purpose | |------|---------| | verify-after-edit.sh | Verify edits worked | | verify-after-write.sh | Verify writes worked | | verify-deployment.sh | Check deployments | | deployment-check.sh | Pre-deploy validation |

Session Management

| Hook | Purpose | |------|---------| | session-persist.sh | Save session state | | session-recovery-check.sh | Recover sessions | | progress-report.sh | Progress tracking | | milestone-tracker.sh | Milestone tracking |

Specialized

| Hook | Purpose | |------|---------| | browser-automation-check.sh | Playwright/Comet checks | | accessibility-check.sh | A11y validation | | performance-alert.sh | Performance monitoring |


Commands

Quick slash commands in commands/ for common actions.

Feature Development

| Command | Purpose | |---------|---------| | /start | Begin new feature | | /feature-dev | Complete feature workshop | | /ship | Complete & create PR | | /idea | Brainstorm & plan |

Code Management

| Command | Purpose | |---------|---------| | /fix | Debug & fix issues | | /test | Run Playwright tests | | /cleanup | Clean up project | | /compare | Compare changes | | /preview | Preview before commit | | /rollback | Undo changes |

Documentation

| Command | Purpose | |---------|---------| | /docs | Generate documentation | | /explain | Explain in plain English | | /status | Project status report |

Session

| Command | Purpose | |---------|---------| | /save-state | Save session checkpoint | | /resume | Resume from checkpoint |

Help

| Command | Purpose | |---------|---------| | /CHEATSHEET | Quick reference | | /COMMAND-MAP | Visual command map | | /GETTING-STARTED | Getting started guide | | /WHICH-COMMAND | Find right command |


Configuration Files

CLAUDE.md (Global Instructions)

The CLAUDE.md file contains global instructions that apply to ALL projects:

  • Agentic Patterns: Reflection, PEV (Plan-Execute-Verify), Metacognitive
  • ID8Pipeline: 12-stage build process for all projects
  • Auto-Approve Commands: Commands that don't need confirmation
  • Comet Browser Automation: Playwright MCP integration

hooks.json

Configures which hooks run and when:

{ "hooks": [ { "event": "pre-commit", "script": "hooks/pre-commit.sh" }, { "event": "post-edit", "script": "hooks/verify-after-edit.sh" } ] }

settings.json

Claude Code preferences (theme, model defaults, etc.)


Project: Reddit Research Agent

Full market research agent at agents/reddit-research/:

cd ~/.claude/agents/reddit-research npm install npx tsx src/agent/index.ts help

Features:

  • Extracts pain points, opportunities, competitor mentions
  • Uses Reddit's hidden JSON API (add /.json to any URL)
  • Web dashboard at localhost:3456
  • Slack/Email alerts
  • Notion sync
  • Multi-agent parallel scanning

File Structure

~/.claude/
├── CLAUDE.md                 # Global instructions
├── README.md                 # This file
├── settings.json             # Preferences
├── hooks.json                # Hook configs
├── mcpServers.json           # MCP servers
│
├── agents/                   # 38+ specialized agents
│   ├── *.md                  # Agent definitions
│   └── reddit-research/      # Full project
│
├── skills/                   # 21 skill modules
│   ├── nextjs-project-manager/
│   ├── supabase-expert/
│   └── ...
│
├── hooks/                    # 26 automation scripts
│   ├── code-review-gate.sh
│   ├── auto-commit.sh
│   └── ...
│
├── commands/                 # 21 slash commands
│   ├── start.md
│   ├── ship.md
│   └── ...
│
└── plugins/                  # Plugin configs
    └── marketplaces/

Tips

Adding a New Agent

  1. Create agents/my-agent.md
  2. Define role, capabilities, and when to use
  3. Commit and push

Adding a New Skill

  1. Create skills/my-skill/ directory
  2. Add skill definition files
  3. Register in manifest if needed

Adding a New Hook

  1. Create hooks/my-hook.sh
  2. Make executable: chmod +x hooks/my-hook.sh
  3. Register in hooks.json

Sync Troubleshooting

Merge conflicts:

cd ~/.claude git stash git pull git stash pop # Resolve conflicts manually

Reset to remote:

cd ~/.claude git fetch origin git reset --hard origin/main

License

MIT - ID8Labs

五维分析
清晰度8/10
创新性6/10
实用性9/10
完整性8/10
可维护性7/10
优缺点分析

优点

  • 促进跨机器同步
  • 支持团队协作
  • 包含多种专业代理和技能

缺点

  • 需要Git知识以有效使用
  • 同步时可能会出现合并冲突
  • 初学者设置可能较复杂

相关技能

ccmp

A
toolCo-Pilot / 辅助式
86/ 100

“看起来很能打,但别让配置把人劝退。”

claude-mods

A
toolCo-Pilot / 辅助式
86/ 100

“看起来很能打,但别让配置把人劝退。”

agentic-qe

A
toolCo-Pilot / 辅助式
86/ 100

“看起来很能打,但别让配置把人劝退。”

免责声明:本内容来源于 GitHub 开源项目,仅供展示和评分分析使用。

版权归原作者所有 eddiebe147.