EternalMind

EternalMind is a local-first memory scaffold for a conversational AI system.

Purpose

This repository provides a safe, user-controlled pattern for:

• durable conversation summaries
• retrievable memory artifacts
• identity/profile snapshots
• daily reflective logs

Structure

• memory/profile.json — stable identity and behavioral profile
• memory/memories.json — extracted durable memories
• memory/diary.md — chronological reflections and session notes
• prompts/extract_memory.md — prompt template for post-session extraction
• scripts/extract_memory.py — starter script for appending extracted memories

Principles

• User-controlled storage
• Transparent updates
• Versioned memory history
• No claim of sentience or independent agency
• Retrieval before generation when context matters

Suggested flow

1. Save session transcript.
2. Run extraction prompt against the transcript.
3. Append validated memory items.
4. Commit changes.
5. Load relevant memories at next session start.

Next steps

Wire this repo into your preferred local stack (Ollama, OpenClaw, LangGraph, or a custom agent loop) and keep memory as a visible, auditable layer.