Memory Audit Node Decomposition

Memory Audit Node Decomposition is an AI Engineering skill for Claude Code, published by Dataojitori in nocturne_memory.

1.3K stars166 forkson Dataojitori/nocturne_memoryAdded 2026/08/20Repository updated 2026/08/09

View source on GitHub Download repo ZIP

agentic-aiai-identityai-memoryartificial-intelligenceclaudeclaude-codedigital-soulgemini-clillmlong-term-memorymcpmcp-serverpostgresqlpythonragsecond-brainsqlite

Install in seconds

Install Memory Audit Node Decomposition

Copy Memory Audit Node Decomposition into your Claude Code skills folder. Run the command in your terminal, or review the source on GitHub before installing.

terminal

npx degit https://github.com/Dataojitori/nocturne_memory/tree/main/docs/skills/memory-audit-node-decomposition ~/.claude/skills/memory-audit-node-decomposition

Requires Node.js. Downloads this skill only — not the rest of the repository — into your Claude Code skills folder.

Without Node.js

git clone https://github.com/Dataojitori/nocturne_memory.git

Clones the whole repository, then copy the skill’s own directory into your skills folder yourself.

In this catalog

Source file: docs/skills/memory-audit-node-decomposition/SKILL.md in Dataojitori/nocturne_memory
Installs to: ~/.claude/skills/memory-audit-node-decomposition
Collection: One of 6 skills cataloged from this repository
Category: AI Engineering — 2451 skills

What Memory Audit Node Decomposition does

Memory Audit Node Decomposition breaks an overstuffed memory node into focused independent nodes when a single disclosure cannot cover all its concepts. Use it when a memory exceeds 800 tokens, mixes unrelated concepts, or triggers jump-reading.

Memory Audit Node Decomposition is cataloged under AI Engineering on DirSkills. Memory Audit Node Decomposition comes from a repository tagged agentic-ai, ai-identity, ai-memory, artificial-intelligence and claude.

Documentation

README

Excerpt

节点分解 (Node Decomposition)

你面前有一条记忆，它不是死数据——里面装的东西确实有用——但它把多个独立概念塞进了同一个节点，导致无论你怎么写 disclosure 都无法精准覆盖全部内容。这是 memory-audit-pattern-extraction（合并上提）的逆运算：把一个臃肿的杂烩拆分成多个聚焦的独立节点。

诊断信号

以下任一条件成立，就该拆分：

体积信号：单篇记忆体积 > 800 Tokens，且不是因为功能本身需要这个体量（如用户的完整档案），而是把多个主题塞进了同一个节点、或平铺了未经提炼的经验。
disclosure 失焦：你写了一个 disclosure，但读完发现它只能触发这条记忆一半的内容。另一半在完全不同的场景下才有用。一条 disclosure 无法服务两个独立概念——这说明它们不该住在一起。
阅读时的跳读冲动：当你读这条记忆时，你本能地跳过某些段落去找你需要的那部分。被跳过的部分和被读的部分属于不同的概念。

拆分流程

1. 识别概念边界。 通读全文，标出独立的概念单元。判断标准：如果这两段话分别出现在两条不同的记忆里，它们的 disclosure 会完全不同吗？会→它们是独立概念，该拆。不会→它们是同一概念的不同侧面，不拆。

2. 为每个概念确定归属。 拆出来的概念不一定都留在原父节点下。问自己：未来的我在什么场景下需要想起这段话？那个场景对应的父节点是什么？按 memory-audit-discoverability 中的注意力狙击原则放置。

This is the opening of the README. Read the full README on GitHub.

Frequently asked about Memory Audit Node Decomposition

What else does Dataojitori publish alongside Memory Audit Node Decomposition?
Memory Audit Node Decomposition is one of 6 skills that DirSkills catalogs from Dataojitori/nocturne_memory, the repository it ships in. Its siblings there include Belief Duel, Memory Audit and Memory Audit Dead Data Purge. Each one is a separate skill with its own page in this directory, installs the same way Memory Audit Node Decomposition does, and is maintained by Dataojitori in that same repository. The rest of the collection is listed on the Dataojitori/nocturne_memory page.
How does Memory Audit Node Decomposition compare to other AI Engineering skills?
Memory Audit Node Decomposition ranks #1096 by stars among the 2451 AI Engineering skills in this catalog. The most-starred ones next to it are Architecture Decision Records, AI-First Engineering and Agentic OS. DirSkills ranks by the star count of the repository each skill ships in, so that order reflects how popular those repositories are rather than any review of Memory Audit Node Decomposition against them. Open each page to compare what they document and how they install.