memory-management

Decode shorthand and acronyms into workplace context using tiered memory files.

1|Updated Feb 25, 2026
One-click install
npx skills add https://github.com/mattmacleod16-svg/freedomforge-max --skill memory-management-mattmacleod16-svg
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-management
Source: https://github.com/mattmacleod16-svg/freedomforge-max/tree/main/.agents/skills/memory-management
Command: npx skills add https://github.com/mattmacleod16-svg/freedomforge-max --skill memory-management-mattmacleod16-svg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Memory management enables Claude to operate as a true workplace collaborator by decoding shorthand, acronyms, nicknames, and internal language so requests are understood as concrete actions.

Core Features & Use Cases

  • Hot cache (CLAUDE.md) surfaces the top ~30 people and terms for quick decoding.
  • Deep memory stores (memory/glossary.md, memory/people/, memory/projects/, memory/context/) scale knowledge as needed.
  • Tiered lookup flow: CLAUDE.md → memory/glossary.md → memory/people/, memory/projects/, memory/context/ to retrieve context.
  • Memory-driven workflows support persistent context across conversations and projects.

Quick Start

Load CLAUDE.md and populate memory/ with core terms and profiles to enable instant decoding.

Frequently Asked Questions about memory-management

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

FAQPage Schema
How do I help Claude decode workplace shorthand and acronyms automatically?

Workplace shorthand and acronyms are decoded automatically by applying a two-tier memory system that maps internal language to actionable context. A hot cache surfaces top terms while deeper memory stores handle extended lookups for consistent retrieval.

What is the best way to maintain contextual memory for Claude across multiple projects?

Contextual memory across projects is maintained by structuring memory into dedicated directories for glossary terms, people, and project details. This allows Claude to retrieve persistent context dynamically during conversations instead of losing prior knowledge.

How does tiered lookup work when retrieving context for Claude?

Tiered lookup retrieves context by first checking a hot cache file for quick decoding of top terms, then cascading to deeper memory stores like glossary and project directories if the requested information is not immediately found.

Do I need to manually populate memory files before using context management?

Yes, you need to manually populate the hot cache and deeper memory stores with your core terms and profiles. Loading these files with workplace vocabulary initially enables the system to instantly decode internal language during interactions.

Can I scale workplace knowledge for Claude without hitting a context limit?

Workplace knowledge scales without hitting limits by moving detailed context into dedicated deeper memory directories. The hot cache retains only the most frequently used items, ensuring quick access while deeper stores manage the bulk of the data.

Why does Claude forget internal nicknames and project details between conversations?

Claude forgets internal nicknames and project details because it lacks persistent memory by default. Implementing a structured memory system with a hot cache and deep memory stores solves this by retrieving saved context on demand.