context-budget

Audit and optimize token overhead across loaded Claude Code session components.

Updated Apr 4, 2026

Search Tags:#performance #optimization #audit #claude #token-savings #context-budget

Ready to run

npx skills add https://github.com/mitul-bhatia/Vibes --skill context-budget-mitul-bhatia

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: context-budget
Source: https://github.com/mitul-bhatia/Vibes/tree/main/.github/skills/context-budget
Command: npx skills add https://github.com/mitul-bhatia/Vibes --skill context-budget-mitul-bhatia

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Audits and optimizes the token usage of a Claude Code session by analyzing every loaded component (agents, skills, MCP servers, rules, and CLAUDE.md) to reclaim context window space and improve performance.

Core Features & Use Cases

Phase 1: Inventory scan all components and estimate tokens, flag heavy items.
Phase 2: Classification categorize components into Always needed, Sometimes needed, or Rarely needed.
Phase 3: Detection identify bloated descriptions, redundant components, and MCP over-subscription.
Phase 4: Reporting generate a Context Budget Report with actionable savings and per-component details.
Use Case: Run /context-budget after adding new skills or MCP servers to ensure there is headroom before deployment.

Quick Start

Use the /context-budget command to generate a full context budget report.

Frequently Asked Questions about context-budget

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

FAQPage Schema

How do I audit token usage and reclaim context window space in a Claude Code session?▼

Run a context budget audit to scan all loaded components, estimate token overhead, and flag heavy items. It categorizes components by necessity and generates a report with actionable savings to reclaim context window headroom.

What is the best way to reduce token overhead from MCP servers and skills?▼

Reducing token overhead from MCP servers and skills requires identifying bloated descriptions, redundant components, and MCP over-subscription. A context budget audit detects these issues and prioritizes optimizations to improve session performance.

Can I check context window headroom before deploying new agents or rules?▼

Yes, you can check context window headroom before deployment by running a context budget audit. It inventories all active agents, skills, rules, and CLAUDE.md files, classifying their token impact to ensure there is enough space.

How does a context budget report classify loaded components for optimization?▼

A context budget report classifies loaded components into Always needed, Sometimes needed, or Rarely needed categories. This classification, combined with per-file heuristics, helps assess impact and prioritize token savings across agents and MCP servers.

Why does my Claude Code session run out of context window space after adding new components?▼

A Claude Code session runs out of context window space when accumulated token overhead from agents, skills, rules, and CLAUDE.md exceeds limits. Auditing identifies bloated descriptions and redundant components to reclaim token headroom.

Does the context budget audit work with custom CLAUDE.md files and rules?▼

Yes, the context budget audit works with custom CLAUDE.md files and rules. It applies token-estimation rules and per-file heuristics across every loaded component to identify headroom, assess impact, and prioritize savings.