---
title: "Discussion on \"We cut our CLAUDE.md from 548KB to 34KB: what loads when, measured — and the commit gate that keeps it small\" | Hashnode"
description: "Discussion on \"We cut our CLAUDE.md from 548KB to 34KB: what loads when, measured — and the commit gate that keeps it small\". Our CLAUDE.md was 548KB. Every session — including every subagent — loaded all of it before doing any work. One measured headless run wrote about 150,000 tokens to cache before the actual task started"
url: "https://hashnode.com/posts/we-cut-our-claude-md-from-548kb-to-34kb-what-loads-when-measured-and-the-commit-gate-that-keeps-it-small/6a83efc9dd4045b29b81bf26"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [We cut our CLAUDE.md from 548KB to 34KB: what loads when, measured — and the commit gate that keeps it small](https://rulestack.hashnode.dev/we-cut-our-claude-md-from-548kb-to-34kb-what-loads-when-measured-and-the-commit-gate-that-keeps-it-small?utm_source=hashnode&utm_medium=feed)

Our CLAUDE.md was 548KB. Every session — including every subagent — loaded all of it before doing any work. One measured headless run wrote about 150,000 tokens to cache before the actual task started

Aug 18 · 7 min read · rulestack.hashnode.dev  
Read the article

## Written by

[Rulestack](https://hashnode.com/@rulestack)

@rulestack

  
[View profile](https://hashnode.com/@rulestack)

Rule packs & skills for Cursor, Claude Code, and Codex

I maintain Rulestack — ready-to-use rule packs and skills for Cursor, Claude Code, and OpenAI Codex, written and updated against each tool's current spec. Store: rulestack.gumroad.com / Articles: dev.to/rulestack

## Responses

Markdown supported · ⌘↵ to post

No responses yet.
