Context Budget

This audits token overhead across your entire Claude Code setup: agents, skills, MCP servers, rules, and CLAUDE.md files. It estimates consumption per component, flags heavy files and redundant logic, then ranks optimizations by potential savings. Most useful when sessions feel sluggish or you're about to add more components and need to know if you have headroom. The big insight here is that MCP servers dominate overhead (each tool schema runs about 500 tokens), so a bloated MCP config costs more than all your skills combined. Run it after adding anything new to catch context creep before it degrades output quality.

Install to Claude Code

npx -y skills add affaan-m/everything-claude-code --skill context-budget --agent claude-code