memory

Consolidate Brain MCP and reflexion memory for hierarchical semantic search and episodic extraction.

Updated Jan 19, 2026

Search Tags:#ai #mcp #memory #episodic-memory #memory-router #reflexion-memory #causal-graphs

Ready to run

npx skills add https://github.com/loriensleafs/brain --skill memory-loriensleafs

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: memory
Source: https://github.com/loriensleafs/brain/tree/main/apps/claude-plugin/skills/memory
Command: npx skills add https://github.com/loriensleafs/brain --skill memory-loriensleafs

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

Memory system for AI agents enabling a memory-first approach: fast semantic search, structured episodic extraction, and causal pattern tracking to ground decisions in past experience.

Core Features & Use Cases

Memory-first semantic search across Brain MCP and Forgetful (Tier 1)
Tier 2 episodic memory extraction from sessions for concise recall
Tier 3 causal graph and pattern tracking to guide decisions and improvements
Real-world use: Before modifying a component, an agent can search past incidents, extract a relevant episode, and apply proven patterns to avoid regressions.

Quick Start

Import and run a basic memory search to surface relevant notes:

Import the memory router module and run a simple search: Import-Module .claude/skills/memory/scripts/MemoryRouter.psm1 $results = Search-Memory -Query "PowerShell arrays" -MaxResults 5

Frequently Asked Questions about memory

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

FAQPage Schema

How do I implement episodic memory extraction for AI agents to recall past sessions?▼

The memory system skill provides Tier 2 episodic memory extraction by structuring session data into concise recall episodes for AI agents. This enables agents to ground decisions in specific past experiences rather than general context.

What is causal pattern tracking for AI memory and how does it work?▼

Causal pattern tracking is a Tier 3 memory feature that builds causal graphs from historical agent actions and outcomes. It guides future decisions by identifying proven patterns to apply and known regressions to avoid during component modifications.

Can I run semantic search across Brain MCP memory without external dependencies?▼

Semantic search operates as a Tier 1 feature directly on local Brain MCP data, requiring no external dependencies. The optional Forgetful endpoint enhances search capabilities but is not required for basic memory-first queries.

Does the memory router module require pre-flight validation before writing causal graphs?▼

The memory router requires pre-flight validation before executing writes to causal graphs to ensure data integrity. This process validates data before updating the hierarchical memory structure across Brain MCP and reflexion memory systems.

When should I use reflexion memory versus standard semantic search for AI agents?▼

Use reflexion memory when agents need structured episodic extraction (Tier 2) or causal pattern tracking (Tier 3) to analyze past incidents and prevent regressions. Standard semantic search (Tier 1) suffices for simple, fast fact retrieval across projects.