🎯Skills4
Audits Claude Code or Codex setups to identify context window waste, then implements fixes to recover 5-25%+ of context capacity. Dispatches parallel agents to audit CLAUDE.md, MEMORY.md, skills, MCP servers, commands, and settings, producing an optimization plan with measurable token savings.
Opens an up-to-date Token Optimizer dashboard showing context usage trends, quality scores, session history, and skill management, with support for both browser-based and file-based viewing in Claude Code and Codex environments.
Audits token waste across multiple agent systems (Claude Code, Codex, OpenClaw) by detecting idle burns, model misrouting, and config bloat with dollar savings estimates.
An interactive context window coaching skill that analyzes your Claude Code or Codex setup, identifies token-efficiency patterns, and provides personalized advice with real measurements for building new projects, optimizing existing setups, or designing multi-agent systems.