itsnex1s

Honed Claude — Documentation skill for Claude Code

Documentation community

Measured, opinionated CLAUDE.md ruleset for Claude Code.

How to install Honed Claude

This entry records only its repository, not the path inside it, so there is no exact command to give. Open itsnex1s/honed-claude and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Honed Claude does

Measured, opinionated CLAUDE.md ruleset for Claude Code. No plugins, no frameworks — one markdown file + benchmarks + research.

Alternatives in Documentation

会话管理（Session Manager）需求文档（PRD / Markdown） — 目标：对 Codex / Claude Code 的本地会话记录进行可视化管理，并提供“一键复制 / 一键终端恢复”能力 31.7k ★
Notion Research Documentation — Research Notion content and synthesize findings into structured briefs 14.6k ★
Repo Recap — Generate a structured recap of the repository state: open PRs, open issues, recent releases, and executive sum 11.9k ★

README

honed-claude

**A measured, opinionated `CLAUDE.md` for Claude Code — no plugins, no frameworks.**

*Refined against 30 days / 30,140 assistant messages of empirical adherence data.*

[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE) [![Claude Code](https://img.shields.io/badge/Claude%20Code-compatible-blueviolet)](https://code.claude.com) [![PRs Welcome](https://img.shields.io/badge/PRs-welcome%20(with%20data)-brightgreen)](#contributing) [![Tokens saved](https://img.shields.io/badge/tokens%20saved-50%25%20per%20session-orange)](benchmarks/measurements.md)

[Install](#install) · [Why](#why) · [Measured Impact](#measured-impact) · [Philosophy](docs/philosophy.md) · [Benchmarks](benchmarks/) · [Research](research/ecosystem-2026.md) · [Token economics](research/token-economics-2026.md)

TL;DR

Anthropic's official guidance: *"if your `CLAUDE.md` is too long, Claude ignores half of it."* This repo is what remains after 30 days of measuring which rules actually change agent behaviour — and cutting the ones that don't.

**One file. 617 tokens per session. Every rule justified by measurement.**

cp CLAUDE.md ~/.claude/CLAUDE.md

That's the whole install.

Why
Measured Impact
Install
What's in the repo
Philosophy
Benchmarks
Research: the OSS ecosystem around AI coding agents
Contributing
Attribution
License

Why

Every plugin you install becomes a dependency you don't control. Every rule you copy from a tutorial is untested against *your* actual agent behaviour.

`honed-claude` is a **single-file, measurement-driven** alternative:

✅ Empirical. Every rule survives an adherence audit against real transcript history.
✅ Minimal. 617 tokens per session — half of what a bloated rulese