allura-memory-skill

Store and retrieve governed AI agent memories via PostgreSQL and Neo4j.

Updated Mar 18, 2026
One-click install
npx skills add https://github.com/Charitablebusinessronin/Allura_Memory --skill allura-memory-skill
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: allura-memory-skill
Source: https://github.com/Charitablebusinessronin/Allura_Memory/tree/main/.opencode/skills/allura-memory-skill
Command: npx skills add https://github.com/Charitablebusinessronin/Allura_Memory --skill allura-memory-skill

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill teaches agents to leverage Allura Brain's governed memory system (Episodic in PostgreSQL and Semantic in Neo4j) via MCP for persistent, auditable memory workflows.

Core Features & Use Cases

  • Store raw memories and curate them into semantic knowledge through a promotion workflow.
  • Retrieve context across project-scoped and global memories for reasoning.
  • Govern memory with policies, audit trails, and versioned updates.

Quick Start

Connect to the memory MCP server and store your first memory to see the governance workflow in action.

Frequently Asked Questions about allura-memory-skill

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

FAQPage Schema
How does governed memory work for AI agents using MCP workflows?

A dual-layer governed memory system stores raw memories in PostgreSQL and promotes them to semantic knowledge in Neo4j, enforcing review policies, audit trails, and strict tenant isolation across projects.

How do I store and promote agent memories to canonical insights?

You store raw memories via the MCP server, then apply strict promotion rules to review and elevate them into canonical semantic knowledge in Neo4j with versioned updates and full audit trails.

What is the best way to manage multi-tenant memory isolation in AI agent databases?

A governed memory architecture with strict tenant isolation ensures project-scoped memories remain separated. The dual-layer PostgreSQL and Neo4j system enforces these boundaries during storage and retrieval.

Do I need both PostgreSQL and Neo4j for governed memory capture and retrieval?

Yes, the dual-layer architecture uses PostgreSQL for episodic raw memory storage and Neo4j for semantic knowledge graphs, enabling the promotion workflow and governed retrieval across projects.

Can I retrieve global and project-scoped memories for agent reasoning across tenants?

Yes, the governed memory system retrieves context across project-scoped and global memories, allowing agents to reason with auditable insights while maintaining strict tenant isolation boundaries.

Why apply memory governance policies instead of standard vector database retrieval?

Memory governance adds promotion rules, versioned updates, and audit trails, ensuring captured memories are reviewed and canonicalized into semantic knowledge rather than stored blindly via standard vector retrieval.