remember

robabby's avatarfrom robabby

Store a memory in the AI-ready vault memory system. Use when encountering information worth preserving across sessions—decisions made, patterns discovered, facts learned, or significant events. Creates properly formatted memory files with metadata.

0stars🔀0forks📁View on GitHub🕐Updated Jan 11, 2026

When & Why to Use This Skill

The 'remember' skill provides a robust long-term memory solution for AI agents by storing structured information in a dedicated, AI-ready vault. It solves the problem of context loss between sessions, enabling the AI to persist critical decisions, learned patterns, and factual knowledge through metadata-rich markdown files that are optimized for future retrieval.

Use Cases

  • Project Continuity: Storing strategic design decisions and rationales to ensure consistent logic and alignment across multiple development sessions.
  • Knowledge Base Building: Capturing semantic facts, definitions, and project-specific details to create a searchable, AI-accessible knowledge repository.
  • Workflow Documentation: Recording procedural patterns and 'how-to' techniques discovered during tasks to optimize future performance and automate repetitive steps.
  • Experience Logging: Tracking episodic breakthroughs, frustrations, or significant events to help the AI adapt its approach based on past interaction history.
nameremember
descriptionStore a memory in the AI-ready vault memory system. Use when encountering information worth preserving across sessions—decisions made, patterns discovered, facts learned, or significant events. Creates properly formatted memory files with metadata.

Remember

Store information in the memory system for cross-session persistence.

Memory Types

Type Folder Use For
Episodic Memory/Episodic/ Events, breakthroughs, frustrations, experiences
Semantic Memory/Semantic/ Facts, definitions, project details, knowledge
Procedural Memory/Procedural/ Patterns, how-tos, workflows, techniques
Strategic Memory/Strategic/ Decisions, approaches, plans, rationale

Workflow

  1. Identify what to remember

    • From $ARGUMENTS if provided
    • Otherwise, ask user what to store
  2. Determine memory type

    • What kind of information is this?
    • Which folder does it belong in?
  3. Assess importance (0.0-1.0)

    • 0.9-1.0: Critical, always relevant
    • 0.7-0.8: Important, frequently useful
    • 0.5-0.6: Moderate, occasionally useful
    • 0.3-0.4: Minor, rarely needed
  4. Extract concepts (keywords for search)

    • 3-6 terms that would help find this later
  5. Create memory file

    • Path: Areas/AI/Memory/{Type}/YYYY-MM-DD - {Brief Title}.md
    • Use frontmatter template below

Memory File Template

---
created: {date}
type: {episodic|semantic|procedural|strategic}
importance: {0.0-1.0}
concepts: [{keywords}]
source: {session|explicit}
---

# {Title}

{Content with sufficient context to be useful standalone}

## Context

{Why this matters, when it applies, related information}

Parameters

  • $ARGUMENTS (optional): The content to remember. If not provided, prompt user.

Default Paths

Memory files are created at: Areas/AI/Memory/{Type}/YYYY-MM-DD - {Brief Title}.md

Related Skills

  • /recall {terms} - Search for stored memories
  • /reflect - Batch memory creation at session end
  • /glean - Surface patterns across memories

Example

User: /remember The decision to use terminal/modern contrast for the AI Ready Vault website

Response: "Storing strategic memory...

Type: Strategic (design decision with rationale) Importance: 0.7 (affects implementation, not critical) Concepts: [ai-ready-vault, design, terminal-aesthetic, visual-hierarchy]

Created: Areas/AI/Memory/Strategic/2025-01-08 - Terminal Modern Contrast Decision.md"