skills.rest
adrianasiam19
adrianasiam19COMMUNITY
Explore all skills by this author →
GitHub

memory-systems

Designs agent memory architectures using vector stores, knowledge graphs, and temporal retrieval.

Updated Jan 31, 2026
Search Tags:#knowledge graph#retrieval#persistence#agent memory#vector store#mem0#temporal reasoning
One-click install
npx skills add https://github.com/adrianasiam19/ATLAS --skill memory-systems-adrianasiam19
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: memory-systems
Source: https://github.com/adrianasiam19/ATLAS/tree/main/.agent/skills/memory-systems
Command: npx skills add https://github.com/adrianasiam19/ATLAS --skill memory-systems-adrianasiam19

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires numpy, and includes scripts (resource) and references (resource) components.

What problem does it solve? AI agents lose all knowledge when sessions end, forcing users to repeat context and preventing long-term personalization. This Skill guides the design and implementation of persistent memory layers so agents retain facts, entities, and relationships across sessions. ## Core Features & Use Cases - Framework Comparison: Compares production memory frameworks (Mem0, Zep/Graphiti, Letta, LangMem, Cognee) with benchmark data from LoCoMo, LongMemEval, DMR, and HotPotQA to support selection decisions. - Layered Memory Architecture: Provides decision tables for working, short-term, long-term, entity, and temporal knowledge graph layers, plus retrieval strategies (semantic, entity-based, temporal, hybrid). - Reference Implementations: Ships runnable Python code for vector stores, property graphs, temporal knowledge graphs, and an integrated memory system combining semantic search with graph traversal. - Use Case: When building a support agent that must remember a customer's preferences and history across conversations, use this Skill to choose a framework, design the memory layers, and implement consolidation logic that invalidates stale facts without discarding history. ## Quick Start Ask the agent to design a memory architecture for a chatbot that must remember user preferences across sessions and recommend whether to use Mem0, Zep, or a simple file-based store.

Frequently Asked Questions about memory-systems

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

FAQPage Schema
How do I add long-term memory to an AI agent?▼

Start with the simplest layer that meets your retrieval needs: file-system storage for prototypes, a vector store like Mem0 for semantic search, or a temporal knowledge graph like Zep/Graphiti for facts that change over time. Add complexity only when retrieval quality degrades.

Mem0 vs Zep vs Cognee: which agent memory framework should I choose?▼

Choose Mem0 for fast time-to-production with managed infrastructure, Zep/Graphiti for bi-temporal modeling and relationship traversal, and Cognee for dense multi-layer semantic graphs with customizable ingestion pipelines. Letta fits agents needing deep self-introspection of their own memory.

What is a temporal knowledge graph for agent memory?▼

A temporal knowledge graph stores facts with validity intervals (valid_from, valid_until) so agents can answer point-in-time queries like where a user lived on a specific date. It prevents stale facts from poisoning context while preserving history.

How do I benchmark an agent memory system?▼

Evaluate against LoCoMo for long-conversation retention, LongMemEval for long-term memory, DMR for dialogue memory, and HotPotQA for multi-hop reasoning. Treat results as signals for specific retrieval dimensions rather than absolute rankings.

Why does my agent retrieve outdated or conflicting memories?▼

Stale memory poisoning happens when facts lack temporal validity tracking or expiry policies. Track valid_from and valid_until timestamps, prefer the most recent fact on conflict, and run periodic consolidation that invalidates old entries without discarding them.

When should I avoid complex memory frameworks?▼

Avoid heavy frameworks during prototyping or for simple agents. Benchmarks show a filesystem-based agent scored 74% on LoCoMo versus Mem0's 68.5%, so reliable basic retrieval often beats sophisticated tooling.

Related Skills

Explore More Software Engineering Skills→
rustamsrustams

context-fundamentals

Explain context engineering concepts for AI agent systems.

Community
Intermediate
LuizEduPPLuizEduPP

memory-systems

Design layered persistence architectures for agent memory frameworks.

Community
Advanced
pagerguildpagerguild

agent-coordination

Coordinates multiple AI agents on parallel tasks while preventing conflicts via a centralized system.

Official
Intermediate
rahulgupta2018rahulgupta2018

memory-systems

Designs persistent memory architectures for agent systems with cross-session retention and entity tracking.

Community
Advanced