skills.rest
Fr-e-d
Fr-e-dCOMMUNITY
Explore all skills by this author →
GitHub

memory-compact

Compress AI agent memory by category and archive originals.

157|28|Updated Feb 7, 2026
Search Tags:#knowledge management#ai agent#memory#compression#context window#token limits
One-click install
npx skills add https://github.com/Fr-e-d/GAAI-framework --skill memory-compact-fr-e-d
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: memory-compact
Source: https://github.com/Fr-e-d/GAAI-framework/tree/main/.gaai/core/skills/cross/memory-compact
Command: npx skills add https://github.com/Fr-e-d/GAAI-framework --skill memory-compact-fr-e-d

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the critical issue of context window limitations by providing an emergency, single-pass method to compress the AI's memory when token limits are being approached during active tasks.

Core Features & Use Cases

  • Emergency Compression: Ideal for immediate memory reduction mid-task without disrupting workflow.
  • Targeted Compaction: Allows for selective compression of memory by category or file size.
  • Durable vs. Ephemeral Classification: Intelligently distinguishes between critical, long-term knowledge and temporary session data, ensuring only appropriate information is archived.
  • Use Case: While working on a complex coding task, you notice the AI is starting to forget earlier instructions due to context window limits. Activating memory-compact will quickly reduce memory usage, allowing the AI to retain focus on the current task.

Quick Start

Activate memory-compact to compress the largest memory category.

Frequently Asked Questions about memory-compact

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

FAQPage Schema
How do I compress AI agent memory when approaching context window token limits?▼

To compress AI agent memory under context window pressure, you can trigger an emergency single-pass memory compression that summarizes selected categories and archives detailed originals to free up tokens.

What is the best way to reduce context window usage mid-task without losing critical knowledge?▼

The best way to reduce context window usage is classifying memory entries by durability, keeping durable knowledge intact while summarizing and archiving ephemeral or superseded session data.

Can I selectively compact AI memory by category or file size?▼

Yes, you can selectively compact AI memory by targeting specific categories or identifying the largest memory files to compress, generating a single summary file for each compacted category.

How does memory compression handle durable versus ephemeral data classification?▼

Memory compression uses an R7 gate to classify entries by durability, preserving critical long-term knowledge while archiving detailed originals only for ephemeral and superseded temporary session data.

When should I use emergency single-pass memory compression for an AI agent?▼

You should use emergency single-pass memory compression when an AI agent begins forgetting earlier instructions during complex tasks due to high context window pressure, requiring immediate token reduction.

Related Skills

Explore More Software Engineering Skills→
bketelsenbketelsen
1

memory-compact

Compress AI memory by classifying entries and generating summary files.

Community
Advanced
delorenjdelorenj
14

letta

Guide Letta development for stateful AI agents with long-term memory.

Community
Advanced
batuhanysvbatuhanysv

context-shredding

Compress runtime context artifacts into full, summarized, or minimal tiers.

Community
Intermediate
Weihong-LiuWeihong-Liu
6

note

Save and manage session-specific notes with priority and working memory options.

Community
Intermediate