creo-memories

Unify a 2-layer architecture with a 4-scene model for persistent cross-session memory.

Updated Dec 20, 2025

Search Tags:#atlas #semantic-search #memory-persistence #context-engine #creo-memories #two-layer-architecture #4-scene-model

Ready to run

npx skills add https://github.com/chronista-club/claude-plugin-creo-memories --skill creo-memories

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: creo-memories
Source: https://github.com/chronista-club/claude-plugin-creo-memories/tree/main/skills/creo-memories
Command: npx skills add https://github.com/chronista-club/claude-plugin-creo-memories --skill creo-memories

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Creo Memories enables persistent context across sessions by providing a robust, 2-layer architecture with a 4-scene model and a 4-cadence self-improvement loop that continuously enhances the ecosystem through automatic context delivery.

Core Features & Use Cases

Deterministic Layer 1 writes (local canon) + Layer 2 cloud traces for dynamic memory state, enabling cross-session continuity and multi-agent collaboration.
4-scene mental model (memories, atlas, views, actions) for structured memory lifecycle, knowledge organization, and actionable workflows.
Semantic search, provenance graphs, and integrated decision recording (ADR style), onboarding, and cycle-close workflows via MCP tooling.

Quick Start

Enable Creo Memories as the default context provider at session start to automatically inject persistent context.

Frequently Asked Questions about creo-memories

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

FAQPage Schema

How do I persist context across sessions for AI agents?▼

To persist context across sessions, you need a memory architecture that unifies local canon writes with cloud traces. This approach enables cross-session continuity and supports multi-agent collaboration through structured memory state management.

What's the best way to organize memory lifecycle for cross-session workflows?▼

Organizing memory lifecycle requires a structured mental model covering memories, atlas, views, and actions. This four-scene framework supports knowledge organization and actionable workflows, ensuring context remains searchable and structured across different operational stages.

How does semantic search work with persistent memory traces?▼

Semantic search over persistent memory traces operates by querying a unified layer of local canon and cloud records. This mechanism retrieves relevant historical context and provenance graphs, enabling accurate information discovery across sessions.

Can I use MCP tooling for integrated decision recording and onboarding?▼

Yes, MCP tooling supports integrated decision recording in ADR style, onboarding, and cycle-close workflows. These tools leverage wedge-documented processes to facilitate collaboration and structured process management within the memory ecosystem.

Does a two-layer architecture improve multi-agent collaboration?▼

A two-layer architecture improves multi-agent collaboration by separating deterministic local canon writes from dynamic cloud traces. This division provides reliable state persistence while allowing agents to share and access evolving context dynamically.

When do I need automatic context delivery in my workflows?▼

You need automatic context delivery when your workflows demand continuous self-improvement and seamless cross-session continuity. A four-cadence loop enhances the ecosystem by injecting persistent context automatically at session start.