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.
npx skills add eranw2000/claude-maintenance-skills --skill split-claude-mdgit clone --depth 1 https://github.com/eranw2000/claude-maintenance-skillsWrote 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.
[](https://agentmods.dev/skills/eranw2000/claude-maintenance-skills/split-claude-md)<a href="https://agentmods.dev/skills/eranw2000/claude-maintenance-skills/split-claude-md"><img src="https://agentmods.dev/badge/skills/eranw2000/claude-maintenance-skills/split-claude-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/eranw2000/claude-maintenance-skills/split-claude-md"><img src="https://agentmods.dev/badge/skills/eranw2000/claude-maintenance-skills/split-claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00187 | $0.03222 |
| Opus 5 | $0.00093 | $0.01611 |
| Sonnet 5 | $0.00037 | $0.00644 |
| Haiku 4.5 | $0.00019 | $0.00322 |
Grade A, and why
split-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 10d 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.
How it starts
The opening of the file, as written. The whole thing — 222 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Split CLAUDE.md - prune, archive, consolidate, extract, compress, scope
An always-loaded file is re-read on every request of every session. Text in it is not stored once; it is paid for continuously. Six things grow over time:
- Old dated records load every turn even when no future session needs them.
- Recurring bugs accumulate as scattered prose with stale instance counts, so a future session cannot tell that a "new" bug is the Nth instance of a known one.
- Stable reference docs sit inline although they are consulted rarely, and every session that does not need them still pays for them.
- Rules accumulate their own evidence, and the war story outgrows the rule.
- Content the repository already answers gets written down and then goes stale, so the file starts contradicting the code.
- Rules that only matter for one file type load in every session regardless.
There are TWO reasons to cut, and the second one is usually missed. Cost is the obvious one. Adherence is the other: a longer file is followed less reliably, which Anthropic states twice. So a file sitting exactly at its cost threshold is not finished, and a rule being ignored is a reason to cut rather than a reason to restate the rule more loudly.
Read this first
reference/loading-model.md holds the mechanics every phase depends on: what
each destination costs, what loads lazily, what is free, what is lost after a
compaction, and the hard limits. Read it before choosing a destination for
anything. Two facts from it decide most of a run:
@pathimports do not reduce context. Imported files load at launch with the file that imports them. This is the usual wrong answer to a large CLAUDE.md and it saves nothing.- A rule moved to a lazy destination is LOST after a compaction until a matching file is read again. A rule that fires on a FILE survives that. A rule that fires on a CONVERSATION does not, and must not be moved.
The six destinations, cheapest last
What ships with it
10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- reference/cross-phase.md 10 KB
- reference/loading-model.md 6.8 KB
- reference/phase-0-prune.md 6.3 KB
- reference/phase-a-archive.md 18 KB
- reference/phase-b-patterns.md 8.4 KB
- reference/phase-c-extract.md 10 KB
- reference/phase-d-rule-evidence.md 4.9 KB
- reference/phase-e-path-scope.md 7.2 KB
- rotate.py 13 KB runs code
- survey.py 33 KB runs code
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.
- 10d ago First seen · 222 lines · 187 tokens per session scan A e207d607b123
split-claude-md is a skill published in the GitHub repository eranw2000/claude-maintenance-skills (5 stars, last pushed 13d ago), licensed MIT. It adds 187 tokens to every session and 3,222 once invoked, about $0.0009 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.
Other skills, from other repositories
resolve-plugins
Curate Claude Code skill / MCP / hook picks against live upstream sources and pin them in .claude/settings.json. Adding a process harness (a paired skill set) is an ordinary pick this door handles. Reads stack from docs/techstack.md, workflow signal from docs/overview.md + existing pins, then live-queries six source…
huggingface-llm-trainer
Train or fine-tune language models with TRL or Unsloth on Hugging Face Jobs, including SFT, DPO, GRPO, reward models, and GGUF conversion. Use for cloud LLM training; use huggingface-vision-trainer for vision tasks.
feature-dev
Guide a feature implementation through a structured seven-phase workflow with deep codebase understanding, clarifying questions, parallel architecture design, and quality review. Use this skill when the user asks to build a new feature, add functionality, or wants a methodical approach to implementation rather than…
memstack-business-gdpr
Use this skill when the user says 'GDPR', 'data protection', 'privacy compliance', 'DPA', 'DSAR', 'data subject request', 'cookie consent', 'privacy audit', 'CCPA', or asks 'do I need GDPR for this repo'. Scans the repository to detect what personal data is collected, classifies sensitivity, determines whether GDPR…
memstack-development-refactor-planner
Use this skill when the user says 'refactor', 'refactoring plan', 'code cleanup', 'reduce duplication', 'simplify code', 'tech debt', 'god class', 'tight coupling', or needs to systematically improve existing code. Identifies targets, assesses risk, and builds incremental execution plans. Do NOT use for writing new…
memstack-product-feedback-analyzer
Use this skill when the user says 'analyze feedback', 'feedback analysis', 'what are customers asking for', or has support tickets, reviews, or survey data to categorize, score, and prioritize into actionable reports. Do NOT use for competitor analysis or market research.