memory-triage

Extract and classify durable conversation facts into long-term memory categories.

62.9k|7.3k|Updated Jun 20, 2023
One-click install
npx skills add https://github.com/mem0ai/mem0 --skill memory-triage
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-triage
Source: https://github.com/mem0ai/mem0/tree/main/openclaw/skills/memory-triage
Command: npx skills add https://github.com/mem0ai/mem0 --skill memory-triage

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a structured approach to extract relevant facts from conversations and store them as durable memories for future personalization and retrieval.

Core Features & Use Cases

  • Memory extraction: Identify and classify facts into categories like identity, preferences, rules, decisions, projects, and relationships.
  • Memory storage & retrieval: Use memory_add to persist memories and memory_search/memory_get to retrieve them across sessions.
  • Lifecycle management: Update, list, and delete memories as needed to keep the memory store accurate.

Quick Start

After each user interaction, extract actionable facts and store them in memory using memory_add with the appropriate category.

Frequently Asked Questions about memory-triage

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

FAQPage Schema
How do I extract durable facts from conversations for long-term memory recall?

You can triage chat interactions to identify durable facts and store them as long-term memory. This process classifies information into categories like identity, preferences, and projects, enabling personalized recall across sessions.

What is the best way to organize long-term memory across different conversation sessions?

The best way to organize long-term memory is to classify facts by identity, preferences, or projects, apply 'As of' dates for temporal anchoring, and use third-person phrasing to ensure accurate retrieval via memory_search.

How do I update or delete stored memories to keep my knowledge base accurate?

You can manage your memory store's lifecycle by using tools to update, list, and delete memories. This ensures your long-term recall data remains accurate and relevant over time.

Can I store sensitive data or credentials when extracting facts from chat history?

No, the memory triage process includes safe handling mechanisms that actively reject credentials and sensitive data, ensuring only safe, durable facts are indexed for long-term recall and personalized responses.

Does memory search support temporal anchoring for personalized responses?

Yes, memory search supports temporal anchoring by attaching 'As of' dates to indexed facts. This allows the system to provide time-aware, personalized responses based on when specific decisions or preferences were recorded.

What categories of information can I classify long-term memories into?

You can classify extracted long-term memories into six core categories: identity, preferences, rules, decisions, projects, and relationships, allowing structured storage and precise future retrieval.