context-budget

Audit Claude Code context window usage across agents, skills, MCP servers, and rules.

2|Updated Mar 17, 2026
One-click install
npx skills add https://github.com/nextc/nextc-claude --skill context-budget-nextc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: context-budget
Source: https://github.com/nextc/nextc-claude/tree/main/nextc-ecc/skills/context-budget
Command: npx skills add https://github.com/nextc/nextc-claude --skill context-budget-nextc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Audits Claude Code context window consumption across agents, skills, MCP servers, and rules. Identifies bloat, redundant components, and produces prioritized token-savings recommendations.

Core Features & Use Cases

  • Inventory across Agents (agents/.md), Skills (skills//SKILL.md), Rules, and MCP servers to quantify token consumption and flag heavy components.
  • Classification and prioritization to surface always-needed vs. on-demand items and generate actionable savings.
  • Use cases include sessions with many skills/agents/MCP servers or planning expansion, where you need clear headroom and optimization strategies.

Quick Start

Run the /context-budget command to generate a context budget report for your current Claude Code session.

Frequently Asked Questions about context-budget

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

FAQPage Schema
How do I audit Claude Code context window token usage?

To audit Claude Code context window token usage, run the /context-budget command. It inventories agents, skills, MCP servers, and rules to quantify consumption and flag heavy components.

Why does my Claude Code session slow down after adding many agents and MCP servers?

Sessions slow down because added agents, skills, and MCP servers consume context window tokens, reducing available headroom. The skill classifies these components to identify bloat and redundant items causing the slowdown.

What is the best way to reduce context waste in Claude Code?

The best way to reduce context waste is to inventory active components and generate prioritized token-savings recommendations. This approach distinguishes always-needed items from on-demand items to target optimizations.

Does this tool identify which MCP servers consume the most tokens?

Yes, it identifies which MCP servers consume the most tokens by performing an inventory across your active servers. It detects issues and produces a final report quantifying overhead per component.