What problem does it solve? MCP servers in claude-flow v3 suffer from slow cold starts (~1.8s), per-request connection overhead, and linear O(n) tool lookups across 213+ tools, making response times too slow for interactive agent workflows. ## Core Features & Use Cases - Connection Pooling: Reuses MCP connections with health checks, idle eviction, and pre-warming to reach 90%+ pool hit rates. - Fast Tool Registry: Builds hash-based O(1) tool indexes with LRU caching and fuzzy matching for sub-5ms lookups. - Load Balancing & Transport Optimization: Distributes requests via least-connections or response-time strategies, with message batching and compression. - Performance Monitoring: Collects p95 latency, error rates, and pool metrics with health status classification. - Use Case: An engineer optimizing a claude-flow v3 deployment uses this Skill to implement all transport-layer improvements and verify the server meets the <100ms p95 response time target. ## Quick Start Ask the AI to analyze the current MCP server performance and implement connection pooling, load balancing, and tool registry optimization to achieve sub-100ms response times.
