context-audit

Audit Claude Code context and produce a health score with prioritized fixes.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/izo/Ulk --skill context-audit-izo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: context-audit
Source: https://github.com/izo/Ulk/tree/main/community-skills/context-audit
Command: npx skills add https://github.com/izo/Ulk --skill context-audit-izo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Audit Claude Code setups for token waste and context bloat. It checks MCP servers, CLAUDE.md rules, skills, settings, and file permissions, and returns a health score between 0-100 with prioritized fixes. Run /context first, then invoke this skill.

Core Features & Use Cases

  • Health scoring: returns a health score 0-100 with prioritized fixes.
  • Multi-area audit: checks MCP servers, CLAUDE.md rules, skills, settings, and permissions.
  • Actionable fixes: provides guidance and steps to improve health.

Quick Start

Run /context to gather the current context, then run /context-audit to generate the health report and prioritized fixes.

Frequently Asked Questions about context-audit

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

FAQPage Schema
How do I audit Claude Code context for token waste and bloat?

You can audit Claude Code context by evaluating MCP servers, CLAUDE.md rules, skills, and settings to identify token waste and context bloat. The audit generates a 0-100 health score and provides prioritized, actionable fixes to optimize your configuration.

What does a Claude Code context health score measure?

A context health score measures the efficiency of your Claude Code setup by evaluating MCP servers, CLAUDE.md rules, skills, settings, and permissions for token waste and misconfigurations. The 0-100 score quantifies overall context health, accompanied by prioritized fixes for any identified issues.

How do I fix Claude Code MCP server and permission misconfigurations?

You fix Claude Code misconfigurations by running an audit that checks MCP servers, settings, and file permissions to detect potential setup errors. The audit outputs a prioritized list of actionable remediation steps to correct token waste and configuration bloat.

Do I need to run /context before auditing my Claude Code configuration?

Yes, you need to run the `/context` command first to gather your current Claude Code environment data before invoking the audit skill. This prerequisite step collects the necessary context information for evaluating token waste and generating an accurate health score.

What are the limitations of auditing Claude Code context for token optimization?

The context audit focuses specifically on typical Claude Code deployments, evaluating MCP servers, CLAUDE.md rules, and settings for token waste and context bloat. It is limited to these defined configuration areas and provides actionable fixes strictly within this scope.