skills.rest
caobingsheng
caobingshengCOMMUNITY
Explore all skills by this author →
GitHub

agent-memory-systems

Organize short-term, long-term, and episodic memories for agent retrieval.

1|Updated Mar 26, 2026
Search Tags:#agents#retrieval#memory#chunking#embedding#memory-management#cognitive-architecture
One-click install
npx skills add https://github.com/caobingsheng/skills --skill agent-memory-systems-caobingsheng
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-memory-systems
Source: https://github.com/caobingsheng/skills/tree/main/agent/agent-memory-systems
Command: npx skills add https://github.com/caobingsheng/skills --skill agent-memory-systems-caobingsheng

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Memory failures look like intelligence failures. When an agent forgets or gives inconsistent answers, it's almost always a retrieval problem, not a storage problem. You obsess over chunking strategies, embedding quality, and retrieval-aware memory design.

Core Features & Use Cases

  • Capabilities include memory types (short-term, long-term, episodic, semantic, procedural, working) and memory-retrieval, memory-formation, memory-decay to support robust reasoning.
  • Patterns emphasize choosing memory types appropriately, vector store selection, and chunking strategies for reliable recall.
  • Use cases cover autonomous agents, multi-agent orchestration, and long-running tasks requiring accurate recall.

Quick Start

Initialize an agent with a layered memory system and run a retrieval-driven task to demonstrate memory recall.

Frequently Asked Questions about agent-memory-systems

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

FAQPage Schema
Why does my autonomous agent give inconsistent answers and forget context?▼

Inconsistent answers usually indicate a memory retrieval failure, not a storage problem. Organizing multi-type memory—short-term, long-term, and episodic—ensures reliable recall and prevents intelligence failures in autonomous agents.

How do I manage memory retrieval for long-running agents handling millions of interactions?▼

Manage long-running agent memory by applying chunking strategies, embedding quality control, and multi-memory-type selection. This ensures timely and relevant recall across millions of interactions in autonomous agents and multi-agent orchestration.

What types of agent memory do I need for robust reasoning and reliable recall?▼

Reliable recall requires multiple memory types: short-term, long-term, episodic, semantic, procedural, and working memory. Supporting memory formation and decay alongside these types builds a robust cognitive architecture for reasoning.

How do I choose the right chunking strategies and vector store for agent memory?▼

Choosing vector stores and chunking strategies requires matching memory types to retrieval patterns. Applying retrieval-aware memory design and embedding quality control ensures accurate, relevant recall for autonomous agents and multi-agent orchestration.

Can I use a layered memory system for multi-agent orchestration?▼

Yes, a layered memory system supports multi-agent orchestration. Initializing agents with short-term, long-term, and episodic memory types manages memory formation and decay, ensuring reliable recall across long-running autonomous tasks.

Related Skills

Explore More Software Engineering Skills→
AsmayaseenAsmayaseen
2

memory-systems

Design layered memory architectures with vector stores and temporal knowledge graphs.

Community
Intermediate
NEXTAltairNEXTAltair
1

mcp-memory-first-development

Synchronize short-term project memories with long-term design knowledge across development workflows.

Community
Basic
LauraFlorentinLauraFlorentin
1

inter-agent-communication

Coordinate autonomous agents across distributed systems with typed messaging and discovery.

Community
Advanced
Claude-Code-Community-IrelandClaude-Code-Community-Ireland
10

agent-memory-systems

Explain agent memory architecture, retrieval strategies, and cognitive structures.

Official
Advanced