memory-never-forget

Name: memory-never-forget
Availability: InStock
Author: kaifashraff

Organize AI-user interactions into temporal and categorical memory layers.

Updated Mar 18, 2026

Search Tags:#personalization #retrieval #memory #long-term #dream-consolidation #temporal-layering

Ready to run

npx skills add https://github.com/kaifashraff/jarvis-research --skill memory-never-forget

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: memory-never-forget
Source: https://github.com/kaifashraff/jarvis-research/tree/main/skills/memory-never-forget
Command: npx skills add https://github.com/kaifashraff/jarvis-research --skill memory-never-forget

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Memory system v3.1 layers memory temporally and categorically to balance recall, relevance, and privacy, automating extraction, consolidation, and drift handling so long-term context remains accurate and useful across sessions.

Core Features & Use Cases

Temporal layering (Sensory, Short-Term, Working, Long-Term) to manage memory lifecycle and decay.
4-type taxonomy (User / Feedback / Project / Reference) for structured, retrievable memories.
Dream consolidation that periodically promotes signals to long-term memory and prunes stale data.
Memory indexing via MEMORY.md with frontmatter-driven metadata to support reliable retrieval and governance.
Memory drift handling to verify current state before applying memories and adapt to changes.

Quick Start

Begin by writing today's daily log to memory/YYYY-MM-DD.md and trigger the nightly Dream consolidation to classify and index memories into MEMORY.md.

Frequently Asked Questions about memory-never-forget

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

FAQPage Schema

How do I implement long-term memory recall for AI conversations across sessions?▼

Temporal memory layering organizes interactions into Sensory, Short-Term, Working, and Long-Term layers. This lifecycle management applies decay over time, promoting only valuable signals to long-term storage while pruning stale data automatically.

What is temporal memory layering and how does it manage context decay?▼

How do I start structuring daily memory logs and triggering dream consolidation?▼

To start structuring daily memory logs, write interactions to a date-based markdown file. Triggering dream consolidation then automatically classifies, indexes, and promotes these memories into a main index using frontmatter-driven metadata.

Does long-term memory retrieval support user personalization and project-specific context?▼

Long-term memory retrieval supports personalization by using a four-type taxonomy that categorizes memories into User, Feedback, Project, and Reference. This structured taxonomy ensures reliable retrieval of specific context across different projects.

How does memory drift handling verify past context before applying it to current sessions?▼

Memory drift handling verifies the current state before applying memories by checking frontmatter-driven metadata. This mechanism adapts to changes over time, ensuring long-term context remains accurate and useful across sessions.

What are the limitations of using markdown files for long-term memory indexing?▼

Using markdown files for long-term memory indexing requires consistent daily logging to trigger dream consolidation. Without regular consolidation, memory drift can occur, causing stale data to persist and reducing reliable retrieval accuracy.