skills.rest
WangK0221
WangK0221COMMUNITY
Explore all skills by this author →
GitHub

Progressive Memory

Organize AI agent memory into daily files and a long-term index with token estimations.

Updated Aug 23, 2026
Search Tags:#context management#information retrieval#ai#memory#token efficiency#progressive disclosure
One-click install
npx skills add https://github.com/WangK0221/OpenClaw_Shrimp --skill progressive-memory-wangk0221
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Progressive Memory
Source: https://github.com/WangK0221/OpenClaw_Shrimp/tree/main/skills/progressive-memory
Command: npx skills add https://github.com/WangK0221/OpenClaw_Shrimp --skill progressive-memory-wangk0221

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the issue of AI context windows being overwhelmed by irrelevant information, leading to wasted attention and decreased performance by implementing a token-efficient memory system.

Core Features & Use Cases

  • Token Efficiency: Significantly reduces token usage by only loading necessary information.
  • Progressive Disclosure: Presents an index first, allowing the AI to fetch detailed information on demand.
  • Structured Memory: Organizes memory into daily files and a long-term memory index with clear categorization and token estimations.
  • Use Case: An AI agent can quickly scan a concise index of past interactions and decisions, then retrieve only the specific details relevant to its current task, such as a past "gotcha" or a critical "rule."

Quick Start

Scan the memory index to find relevant information for the current task.

Frequently Asked Questions about Progressive Memory

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

FAQPage Schema
How do I reduce token usage when loading AI memory into the context window?▼

Progressive disclosure addresses AI context limits by organizing memory into daily files and a categorized long-term index, allowing the agent to scan summaries and fetch specific details only when needed.

How does progressive disclosure work for AI agent memory retrieval?▼

Progressive disclosure presents a categorized long-term memory index with token estimations first. The AI agent scans this index to locate relevant past interactions, then retrieves only the specific daily file details required for its current task.

How do I organize long-term memory files to optimize AI context management?▼

Organize long-term memory by structuring data into daily files and a categorized index containing token estimations. This allows the AI to prioritize scanning the index over loading full memory files, optimizing context window usage.

Can I fetch specific past interactions on demand without loading the entire memory history?▼

Yes, you can fetch specific past interactions on demand by scanning a structured memory index. The system estimates token sizes for categorized entries, allowing the AI to retrieve only the relevant daily file details for its current task.

When should I use a token-efficient memory system over loading full context for AI agents?▼

Use a token-efficient memory system when AI context windows are overwhelmed by irrelevant information, causing wasted attention and decreased performance. It is ideal when you need to quickly reference past decisions or rules without loading full memory histories.

Related Skills

Explore More Software Engineering Skills→
William2333ZZWilliam2333ZZ
5

memory

Store and retrieve persistent workspace memory in MEMORY.md and daily logs.

Community
Intermediate
Fr-e-dFr-e-d
157

memory-compact

Compress AI agent memory by category and archive originals.

Community
Intermediate
vrcmsvrcms
1

token-budget-advisor

Estimate token usage and control response depth for AI queries.

Community
Intermediate
agilebydesignagilebydesign
1

abd-embed-vectors

Embed Markdown text chunks into a local FAISS vector index via the OpenAI API.

Official
Intermediate