Multi-Agent Collaboration

关于

Multi-Agent Collaboration provides state-of-the-art architectural patterns for building complex, context-aware agentic workflows. By implementing strategies like the Four Laws of Context Management and tiered context fidelity, this skill helps developers reduce context overhead by up to 78% while maintaining high analysis quality. It guides users through selecting the optimal orchestration pattern—whether hierarchical, swarm, or ReAct—ensuring Claude-powered systems remain performant, cost-effective, and scalable by avoiding common anti-patterns like snapshot broadcasting.

主要功能

• 0 GitHub stars
• Progressive disclosure techniques to minimize initial prompt overhead
• Tiered context fidelity models for efficient role-based data distribution
• Architectural selection framework including Hierarchical, Swarm, and ReAct patterns
• Context engineering patterns achieving 78%+ reduction in token usage
• Standardized handoff protocols for seamless agent-to-agent communication

使用场景

• Implementing security-focused agent swarms with selective data projection
• Debugging performance bottlenecks and context issues in multi-agent systems
• Designing complex autonomous agent workflows with minimal token overhead