claude-md

claude-md is a skill for Claude Code from fullymiddleaged/Clawness. It costs 91 tokens per session (2,104 once invoked), scanned A, original, MIT.

A guided process for shortening an oversized CLAUDE.md file, the project instruction file that Claude Code loads in every session.

In plain words
What is it for?
Reviewing and reorganising CLAUDE.md, measuring its session cost, removing information already documented elsewhere, and separating path-specific or prompt-specific rules.
Why use it?
Large always-loaded instructions use context and can make important guidance harder for Claude Code to follow. The skill moves specialised instructions into separate rule files after checking their replacements.

Skill for Claude Code

Written for Claude Code: SessionStart hook event. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions Claude Code.

Part of the clawness plugin — 13 skills, 7 agents shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/fullymiddleaged/clawness/claude-md
Any agent
npx skills add fullymiddleaged/Clawness --skill claude-md
Clone the repo
git clone --depth 1 https://github.com/fullymiddleaged/Clawness

Made for: Claude Code.

Or install clawness, the plugin that ships this one along with the rest of its 13 skills, 7 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for claude-md

README.md
[![agentmods](https://agentmods.dev/badge/skills/fullymiddleaged/clawness/claude-md.svg)](https://agentmods.dev/skills/fullymiddleaged/clawness/claude-md)
Your own site
<a href="https://agentmods.dev/skills/fullymiddleaged/clawness/claude-md"><img src="https://agentmods.dev/badge/skills/fullymiddleaged/clawness/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 91 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,104 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00091 $0.02104
Opus 5 $0.00046 $0.01052
Sonnet 5 $0.00018 $0.00421
Haiku 4.5 $0.00009 $0.00210

Measured 5d ago against content hash 1215ac232505, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

claude-md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/claude-md/SKILL.md · 155 lines

How it starts

The opening of the file, as written. The whole thing — 155 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Shrink CLAUDE.md

CLAUDE.md is loaded by the harness, in full, at the start of every session and re-read after /compact. It is the one context cost nothing can cap: Clawness's rules block is budgeted, the lessons log is ranked and budgeted, CLAUDE.md is neither. Files past ~200 lines also measurably reduce how reliably Claude follows any single instruction in them, so a big CLAUDE.md costs twice — tokens, and the adherence of the rules you actually cared about.

This skill is the remedy. The Clawness SessionStart check is only the diagnosis: it reports the number once and deliberately refuses to start this work, because it fires before you've said what you came for. You invoking this skill is the consent that note couldn't ask for.

Say this before starting. This is a destructive reorganisation of a file the user may care about a great deal, and it takes a while. Tell them: it will take most of a session, nothing is deleted until its replacement is verified, and they can stop after any single section. Then get an explicit go-ahead.

First: is this the right tool?

Claude Code ships /doctor (alias /checkup, v2.1.206+), which proposes CLAUDE.md trims natively and migrates what remains into skills and nested CLAUDE.md files. It reports before changing anything.

  • If the project does not use Clawness project rules, say so and recommend /doctor instead. It is maintained by the harness and knows more about how instructions load than this skill does. Do not compete with it out of loyalty.
  • This skill earns its place when the project uses Clawness, because it has two destinations /doctor doesn't know exist: .clawness/rules/ (retrieved by relevance and budgeted, with no line cap and the full rule format) and .clawness/memory.md (one-line traps). Running both is fine and they compose — /doctor for the trim, this for the Clawness-specific placement.

What actually reduces per-turn cost

Only two harness mechanisms do, plus Clawness's own retrieval. Get this right or the whole exercise is theatre:

Read the full file on GitHub · 155 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 155 lines · 91 tokens per session scan A 1215ac232505

Subscribe to this mod's changes

claude-md is a skill published in the GitHub repository fullymiddleaged/Clawness (3 stars, last pushed 7d ago), licensed MIT. It adds 91 tokens to every session and 2,104 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

linked-intent-dev

Guide for linked-intent development (LID). Consult for ALL code changes. Walks changes through a mode-aware six-phase workflow (HLD → LLD → EARS → intent-narrowing edge audit → tests-first → code) with mandatory stops between each phase. Bugs walk the arrow like any other change — no short-circuit. Enforces cascade…

jszmajda/lid · 87 tokens

update-lid

Configure or reconcile a project for linked-intent development (LID). Dispatches on project state — fresh bootstrap, append directives to an existing agent-instructions file (AGENTS.md or CLAUDE.md), add missing mode marker, reconcile convention drift, or run mode transitions. Invoked as /update-lid. For fresh…

jszmajda/lid · 105 tokens

arrow-maintenance

Navigation and audit overlay for linked-intent development. Use when working with docs/arrows/ — orienting via index.yaml, auditing spec-to-code coherence, detecting reverse orphans and drift, splitting/merging/renaming/re-parenting segments. Dual-mode: ambient guidance when the overlay is present…

jszmajda/lid · 85 tokens

map-codebase

Bootstrap LID in an existing (brownfield) codebase. Deep-reads every file in the declared scope, offers lens-based clustering options, generates skeleton LLDs/HLD/EARS bottom-up, then creates arrow docs and prompts the user to flesh out the skeletons. Token-intensive by design. Use when asked to map a codebase…

jszmajda/lid · 92 tokens

bidirectional-differential

Audit coherence across an arrow of intent by running two parallel fresh Claude sessions — one reconstructs code from a single EARS, the other reconstructs the EARS from stripped code — then classifies the drift between them. Use when the user invokes /differential-audit, asks to audit EARS-to-code drift for a feature…

jszmajda/lid · 172 tokens

lid-coach

Review a project's current linked-intent-development (LID) usage against LID's own principles and produce a prioritized report of recommendations for getting more out of the methodology. Invoke when the user runs /lid-coach, asks for a LID review, asks how they could use LID better, wants feedback on their LID setup…

jszmajda/lid · 117 tokens