Context optimization for AI coding agents — lower token cost and latency without sacrificing accuracy.

Cut context bloat in your AI-agent stack: find and safely prune unused skills, MCP servers and subagents from real transcript evidence