Airsc miễn phí cho tất cả. Nếu nó giúp bạn, hãy ủng hộ $2–$20.

Ủng hộ
AgentCẩn thậnQuy trình & bộ nhớ agentCần cài đặt chút

claude-007-agents

avivl · avivl/claude-007-agents

Hệ thống này điều phối các AI agent chuyên biệt cho Claude Code trong nhiều tác vụ phát triển, cải thiện quy trình làm việc với khả năng phối hợp, độ bền và bộ nhớ tổ chức.

🤖 A unified AI agent orchestration system featuring 10's of specialized agents across 14 categories for modern software development. Built with advanced coordination intelligence, resilience engineering, and structured logging capabilities.

270 sao32Giấy phép: MITCập nhật năm ngoái0 lượt mở

Xem trước

▍

Bạn có thể làm gì với nó

  • Xây dựng một hệ thống xác thực bền vững bằng cách phối hợp các agent bảo mật và backend.
  • Phát triển một nền tảng thương mại điện tử bằng cách tự động quản lý nhiều agent chuyên biệt.
  • Tận dụng bộ nhớ tổ chức để tái sử dụng các mẫu và ngữ cảnh từ các dự án trước.

Kiểm tra an toàn

Cẩn thận

Dùng được, nhưng nên đọc ghi chú trước.

  • ●Hơn một năm không cập nhật.

Quét tự động các file trong repo. Luôn xem lại code trước khi chạy.

Trích từ README

🚀 Claude Code: Comprehensive AI Agent System for Developers

Enhance your development workflow with specialized AI agents that help deliver quality code through orchestration, resilience engineering, and organizational memory.

Open-source AI agents for enhanced development workflows

🌟 Why Claude Code Agents Can Improve Your Development

⚡ Specialized Expertise Across Technologies

Reduce context switching between documentation, Stack Overflow, and your IDE. Get access to specialized agents covering major frameworks, languages, and domains.

# Instead of spending hours researching...
claude "Build a resilient authentication system with circuit breakers and structured logging"
# → @rails-expert + @resilience-engineer + @security-specialist work in perfect coordination

🧠 AI System with Organizational Memory

These agents can remember and learn from projects when configured with Basic Memory MCP, helping build institutional knowledge over time.

  • Pattern Reuse: "Use patterns from previous authentication implementations"
  • Context Building: "Reference similar microservice architectures we've built"
  • Knowledge Evolution: "Update our React component patterns based on this success"

🎭 Agent Orchestration System

The system includes coordination features that can help with agent selection and workflow management for complex tasks.

# Just describe what you want - the system handles everything automatically:
claude "Build an e-commerce platform"
# → Vibe Coding Coordinator: 15-20 minute preparation phase, analyzes requirements
# → Exponential Planner: Strategic planning with capability scaling awareness
# → Parallel Coordinator: Coordinates (@react-expert + @rails-expert + @payment-specialist + @security-specialist)
# → Session Manager: Maintains context across extended development sessions
# → Safety Specialists: Pre-deployment validation and architectural analysis

🚀 IMPORTANT: Bootstrap Required ⚠️

⚡ ONE-TIME SETUP: Every project needs bootstrapping before using Claude 007 Agents!

# After installing agents, bootstrap each project once:
claude "Use @bootstrap-orchestrator to analyze and setup this project"

Why Bootstrap?

  • 🔍 Analyzes your tech stack and selects optimal agents
  • ⚙️ Creates proper CLAUDE.md configuration with commit attribution
  • 🎯 Sets up project-specific agent selection and workflows
  • 📋 Configures Task Master integration for codebase-aware development
  • ✅ Validates complete system readiness

Skip this step and agents won't work properly in your project!

🚀 Key Features

1. 🏗️ Engineering Excellence by Design

Every agent prioritizes production-grade quality:

  • ⚡ Fault Tolerance: Circuit breakers, retry mechanisms, graceful degradation
  • 📊 Structured Logging: JSON logging with contextual information
  • 🔄 Self-Healing: Automatic recovery patterns
  • 🛡️ Automated Quality: Trunk.io integration for linting and security

2. 🤖 Advanced Orchestration System

Built on insights from Anthropic's "Code with Claude" conference:

  • Vibe Coding Coordinator: 15-20 minute preparation phases for autonomous development
  • Exponential Planner: Strategic planning with AI capability doubling awareness (7-month cycles)
  • Session Manager: Resumable workflows with context preservation across extended sessions
  • Parallel Coordinator: Multi-agent execution with coordinated tool calling
  • Safety Specialists: Pre-deployment testing, architectural analysis, and risk management
  • Sequential Thinking: Complex multi-step reasoning with adaptive planning and course correction

3. 🚀 Task Master Integration ⭐ NEW

Revolutionary codebase-aware autonomous development ecosystem:

  • Codebase-Aware Intelligence: Tasks generated with deep understanding of existing code architecture and patterns
  • Bridge Agent Architecture: 5 specialized Task Master agents seamlessly integrated with Claude 007's 112 agents
  • Multi-Model AI Configuration: Support for Claude, Gemini, Perplexity, OpenAI, and more
  • Template Intelligence: Smart generation of MCP configs, environment files, and PRD templates
  • Autonomous Development: 30-40% development time reduction with quality assurance

4. 🎯 Live MCP Integrations

Direct access to live systems and real-time data:

  • GitHub MCP: Live repository operations and PR management
  • Task Master MCP: Project management with complexity analysis and codebase awareness
  • Context7 MCP: Up-to-date library documentation
  • Basic Memory MCP: Persistent organizational knowledge
  • Sequential Thinking MCP: Complex reasoning with adaptive planning and course correction

5. 📋 Professional-Grade Workflows

  • Vibe Coding: 15-20 minute autonomous development preparation phases
  • Parallel Execution: Coordinated multi-agent development with tool call batching
  • Exponential Planning: Long-term development planning with AI capability scaling awareness
  • Sequential Reasoning: Multi-step problem analysis with adaptive strategy revision
  • Safety Validation: Pre-deployment testing and architectural safety analysis
  • Dynamic Permissions: Real-time permission management with risk-based escalation
  • Automatic Commit Attribution: Every commit traces back to contributing agents
  • Quality Gates: Pre-commit hooks with comprehensive linting

6. 🧠 Organizational Intelligence

  • Cross-Project Learning: Patterns and knowledge shared across all projects
  • Decision History: Track architectural decisions and their outcomes
  • Performance Analytics: Agent effectiveness measurement and optimization
  • Knowledge Graphs: Dynamic project understanding with real-time updates

🎯 Perfect For

Team SizeUse CaseKey Benefits
Solo DevelopersFull-stack
Xem mã nguồn trên GitHub