Does this skill require an external database?

No, it operates entirely on local Markdown files within your repository, making it transparent, portable, and easy to search using standard tools like Grep.

Can the AI change my core values automatically?

No. For safety and accuracy, the L4 Core Layer is read-only for the AI. Claude can suggest updates or generate content for you to review, but you must manually update the file yourself.

How does the skill detect recurring patterns?

When a preference or behavior is mentioned, the skill uses Grep to scan previous L1 records. If the pattern appears three or more times, it suggests moving that data to the L2 Behavioral layer.

What are the four levels of the memory system?

The system organizes data into L1 (Situational/Daily Events), L2 (Behavior/Habits), L3 (Cognitive/Thinking Models), and L4 (Core Values/Beliefs).

Personal Memory Recorder

Name: Personal Memory Recorder
Author: zephyrwang6

byzephyrwang6

•

생산성 및 워크플로

Automates the capture and categorization of key insights, decisions, and preferences into a structured, four-tier hierarchical memory system.

소개

The mem-record skill provides Claude with a sophisticated personal memory architecture that organizes information into four distinct layers: L1 (Context), L2 (Behavior), L3 (Cognition), and L4 (Core Values). It monitors conversations for explicit requests or implicit signals like major decisions and habit patterns, then automatically extracts relevant data to update local Markdown-based logs. This system enables long-term context retention without a vector database, using pattern detection to suggest elevating recurring behaviors into behavioral habits or core cognitive frameworks.

주요 기능

Automated Markdown file management for long-term storage
Four-tier hierarchical memory categorization (L1-L4)
Read-only protection for L4 core values with suggestion-only updates
37 GitHub stars
Pattern detection for recurring behaviors via Grep-based analysis
Automatic extraction of events, decisions, preferences, and emotions

사용 사례

Tracking recurring UI/UX preferences to build a personalized design workflow
Building a long-term personal knowledge base through daily conversation logs
Documenting project architectural decisions and their underlying rationale

소개

주요 기능

Automated Markdown file management for long-term storage
Four-tier hierarchical memory categorization (L1-L4)
Read-only protection for L4 core values with suggestion-only updates
37 GitHub stars
Pattern detection for recurring behaviors via Grep-based analysis
Automatic extraction of events, decisions, preferences, and emotions

사용 사례

Tracking recurring UI/UX preferences to build a personalized design workflow
Building a long-term personal knowledge base through daily conversation logs
Documenting project architectural decisions and their underlying rationale