memory-management

Decode shorthand, acronyms, and nicknames into structured working memory for Claude.

1|Updated Mar 9, 2026
One-click install
npx skills add https://github.com/kiryteo/opencode-setup --skill memory-management-kiryteo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-management
Source: https://github.com/kiryteo/opencode-setup/tree/main/skills/memory-management
Command: npx skills add https://github.com/kiryteo/opencode-setup --skill memory-management-kiryteo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Memory management provides a structured two-tier memory layer for Claude, enabling rapid decoding of internal language, acronyms, and nicknames to actionable context within a workplace.

Core Features & Use Cases

  • Two-tier architecture: hot cache (CLAUDE.md) for frequent terms and deep memory (memory/ glossary, people, projects, context) for long-tail knowledge.
  • Tiered lookup workflow: CLAUDE.md → memory/glossary.md → memory/people/ & memory/projects → memory/context for comprehensive understanding.
  • Rich interaction patterns: supports onboarding, memory updates, and progressive disclosure to scale memory without cluttering the hot cache.

Quick Start

Populate memory with Todd Martinez as a contact and set his communication preferences to Slack.

Frequently Asked Questions about memory-management

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

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

To decode workplace shorthand, this skill uses a two-tier memory system that translates internal acronyms and nicknames into actionable context, storing frequent terms in a hot cache and long-tail knowledge in deep memory directories.

How do I structure long-tail project context for AI workflows?

To structure project context, use deep memory directories including a glossary file for terms, separate folders for people and projects, and a context directory, ensuring comprehensive understanding through a tiered lookup workflow.

What is a two-tier memory architecture for AI agents?

A two-tier memory architecture combines a hot cache for frequently accessed terms with deep memory storage for long-tail knowledge, enabling rapid decoding of internal language without cluttering the primary working context.

How do I populate AI memory with contact profiles and communication preferences?

To populate memory with contact profiles, add entries to the people directory including their names, nicknames, and specific communication preferences like Slack, which the system then decodes during daily task execution.

When should I use a hot cache versus deep memory directories?

Use a hot cache for frequent terms requiring immediate access, and use deep memory directories for long-tail knowledge, applying progressive disclosure to scale memory storage without overwhelming the working context.

Do I need frontmatter requirements to manage AI memory workflows?

Yes, managing memory workflows requires frontmatter with name and description fields in the skill definition file, enforcing structured metadata to support the tiered lookup process and memory updates.