skills.rest
nsalvacao
nsalvacaoCOMMUNITY
Explore all skills by this author →
GitHub

memory-management

Decode shorthand and internal terms using CLAUDE.md and memory/ files.

Updated Feb 9, 2026
Search Tags:#context#claude#glossary#knowledge-base#lookup#memory-management#decoding
One-click install
npx skills add https://github.com/nsalvacao/nsalvacao-claude-code-plugins --skill memory-management-nsalvacao
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: memory-management
Source: https://github.com/nsalvacao/nsalvacao-claude-code-plugins/tree/main/plugins/productivity/skills/memory-management
Command: npx skills add https://github.com/nsalvacao/nsalvacao-claude-code-plugins --skill memory-management-nsalvacao

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Turns shorthand and internal terminology into actionable Claude context, reducing misinterpretation and delays in work.

Core Features & Use Cases

  • Two-tier memory architecture: hot cache in CLAUDE.md plus a deep memory store in memory/, to balance speed and depth.
  • Tiered lookup and decoding: decode common terms in CLAUDE.md, then consult memory/glossary, memory/people, memory/projects, and memory/context for full context.
  • Rich context when needed: attach profiles, project details, and company context to assist execution on complex tasks.

Quick Start

Initialize CLAUDE.md as the hot cache and load memory/ for full context to enable fast, accurate decoding of internal terms.

Frequently Asked Questions about memory-management

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make Claude understand internal company shorthand and terminology?▼

To make Claude understand internal company shorthand, you use a two-tier memory architecture that decodes terms by checking a hot cache in CLAUDE.md before consulting deep memory files like memory/glossary for full context.

How does the two-tier memory architecture work for maintaining project context?▼

The two-tier memory architecture works by keeping a hot cache in CLAUDE.md for fast term decoding, while a deep memory store in memory/ directories holds detailed profiles, project details, and company context for complex tasks.

Can I use local files to build a knowledge base for Claude without external dependencies?▼

Yes, you can use local files to build a knowledge base without external dependencies. The system loads terms, profiles, and contexts directly from local memory files using a tiered lookup flow.

What is the best way to structure a glossary lookup flow for an AI assistant?▼

The best way to structure a glossary lookup flow is a tiered approach: decode common terms in a CLAUDE.md hot cache first, then consult deep memory directories like memory/glossary, memory/people, and memory/projects for full context.

Do I need external APIs to maintain deep memory and project context in Claude?▼

No, you do not need external APIs to maintain deep memory and project context. The architecture relies entirely on local files, loading terms and contexts with no external dependencies required.

Related Skills

Explore More Content & Communication Skills→
markus41markus41
1

authentication

Implement JWT, OAuth2, sessions, and RBAC for authentication and authorization.

Community
Intermediate
chelleboyerchelleboyer

memory-management

Decode shorthand and acronyms using CLAUDE.md and memory/ tiered lookup.

Community
Advanced
anthropicsanthropics
23.4k

memory-management

Decode workplace shorthand and acronyms using a two-tier memory system.

Official
Intermediate
coco-researchcoco-research
264

brain

Manage a SQLite-backed project knowledge base for structured memory.

Community
Advanced