mem-record

Extract events, decisions, preferences, and emotions from conversations into layered memory levels.

333|83|Updated Jan 9, 2026
One-click install
npx skills add https://github.com/zephyrwang6/myskill --skill mem-record
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mem-record
Source: https://github.com/zephyrwang6/myskill/tree/main/mem-record
Command: npx skills add https://github.com/zephyrwang6/myskill --skill mem-record

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

AI personal memory systems often struggle to capture and organize key information from conversations. This Skill automatically extracts events, decisions, preferences, and emotions from dialogue and stores them into the appropriate memory layers (L1-L3, with L4 guidance for manual review).

Core Features & Use Cases

  • Automatic extraction: Pulls events, decisions, preferences, emotions, and actions from conversations.
  • Layered storage: Routes insights to L1_情境层, L2_行为层, and L3_认知层, with prompts for L4核心层 when needed.
  • Pattern detection: Detects repeated motifs to suggest higher-level summaries.
  • Traceability: Keeps timestamps and sources for later review and analytics.

Quick Start

Initiate by saying 记录到记忆系统 to trigger extraction and layering of the current dialogue into the memory hierarchy.

Frequently Asked Questions about mem-record

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

FAQPage Schema
How do I automatically extract and organize information from conversations into memory layers?

Automatically extract key information from conversations and organize it into structured memory layers (L1-L3) by triggering the memory system. The Skill routes events, decisions, preferences, and emotions to appropriate layers, keeping timestamps and sources for traceability.

What memory layers does the system use and how are conversations routed?

The system uses four memory layers: L1_情境层 (context), L2_行为层 (behavior), L3_认知层 (cognition), and L4核心层 (core, for manual review). Conversations are automatically routed to the appropriate layer based on content type, with pattern detection flagging repeated motifs for higher-level summaries.

Can I use structured memory automation for daily chats, decisions, and preferences?

Yes. The Skill captures events, decisions, preferences, and emotions from daily conversations and stores them into the memory hierarchy automatically. It supports frontmatter-driven discovery and optional scripts for extensibility.

How do I trigger memory extraction from a conversation?

Initiate memory extraction by saying 记录到记忆系统 (record to memory system) to trigger automatic extraction and layering of the current dialogue into the memory hierarchy with timestamps and source tracking.

Does the system detect repeated patterns across conversations?

Yes. Pattern detection identifies repeated motifs across conversations and flags them to suggest higher-level summaries. This enables refinement of memory organization and insights over time.