token-hygiene

Audit Claude Code system context token overhead across CLAUDE.md, MEMORY.md, skills, MCP servers, and logs.

1|Updated Mar 5, 2026
One-click install
npx skills add https://github.com/AlfonsoSBLA/claude-token-hygiene --skill token-hygiene-alfonsosbla
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: token-hygiene
Source: https://github.com/AlfonsoSBLA/claude-token-hygiene/tree/main
Command: npx skills add https://github.com/AlfonsoSBLA/claude-token-hygiene --skill token-hygiene-alfonsosbla

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill audits Claude Code system token overhead across CLAUDE.md, MEMORY.md, installed skills, MCP servers, and daily logs, helping teams reduce overhead and maintain snappy conversations.

Core Features & Use Cases

  • Measures and reports per-conversation token overhead across system context to highlight bloats.
  • Produces a repeatable monthly report with actionable cleanup recommendations for CLAUDE.md, MEMORY.md, skill listings, and MCP configurations.
  • Use cases include monthly housekeeping, on-demand audits when performance slows, and baseline tracking over time.

Quick Start

Run the token hygiene skill to generate the monthly overhead report.

Frequently Asked Questions about token-hygiene

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I reduce Claude Code token overhead in my project context?

Audit Claude Code token overhead by analyzing CLAUDE.md, MEMORY.md, skill descriptions, MCP servers, and daily logs to identify bloats and prioritize actionable optimizations for your project context.

What is the best way to track system context token waste over time?

Generate a reproducible monthly report that measures per-conversation token overhead, enforcing thresholds for file sizes, lines, and server counts to track system context waste over time.

How do I audit CLAUDE.md and MEMORY.md files for bloat?

Audit CLAUDE.md and MEMORY.md by applying strict thresholds for file sizes and line counts, producing a structured report with cleanup recommendations to eliminate configuration bloat.

Does the token hygiene audit work without external dependencies?

The token hygiene audit runs without external dependencies, requiring only your existing Claude Code system files and MCP configurations to quantify overhead and recommend actions.

When should I run a token overhead audit on my MCP servers?

Run a token overhead audit during monthly housekeeping or on-demand when conversations slow down, checking MCP server counts against enforced thresholds to identify configuration bloats.