Skip to content

Repository files navigation

🧠 The DIY Nervous System

A Manual RAG Toolkit for Long-Term AI Memory

"Stop using your AI as a disposable calculator. Start building a second brain."


1. The Problem: The Amnesia Loop

You have brilliant conversations with your AI (ChatGPT, Claude, Gemini). You build context, define rules, and reach deep insights. Then the session ends. Or the context window fills up. And everything is gone.

Most users treat AI memory like a bucket: You fill it until it overflows, then you start empty. This Toolkit allows you to treat AI memory like a Nervous System: A structured, external network of knowledge that grows stronger over time, not weaker.

2. The Solution: Vectorized Manual Indexing

We don't need complex vector databases or Python scripts running on servers to fix this. We just need Structure.

This Toolkit provides you with 4 System Prompts (AI Personas) that allow you to build a sophisticated, navigable knowledge base using nothing but text files and your current AI chat.

It turns your chaotic folder of screenshots and notes into a "Reef"—a self-organizing ecosystem of knowledge.


3. The Toolchain (The 4 Agents)

You don't run code. You paste these prompts into your chat to activate specialized agents.

🏛️ 1. The Container: Master_Index_Template.md

The empty skeleton of your new nervous system. It defines 5 Habitats (Law, History, Ideas, Mechanics, Reference) where your knowledge will live.

👁️ 2. The Optical Alchemist (v3.0)

  • Input: Screenshots, PDF pages, Diagrams.
  • Function: High-fidelity OCR and Semantic Analysis.
  • Output: Clean, standardized JSON files ("Knowledge Assets").
  • Why: Turns dead pixels into machine-readable fluid text.

🧭 3. The Astrolabe (v3.0)

  • Input: The JSON files from the Alchemist.
  • Function: Calculates the "Vector Coordinates" of the information.
  • Output: A 3-line navigation block for your Index.
  • Why: Determines where a file belongs and how the AI should treat it (is it immutable Law? Or just a brainstorming Idea?).

🧹 4. The Index Curator (v3.0)

  • Input: The loose blocks from the Astrolabe.
  • Function: Sorts and files everything into the Master Index without breaking the links.
  • Why: Keeps your system clean automatically.

4. The Physics (How the AI navigates)

This system works because it teaches the AI to read Metadata Vectors instead of just keywords.

Vector Symbol Meaning
Alignment ψA "Trust Score". Is this file core to your mission (0.99) or just a random reference (0.50)?
Density ζ "Compression". Is this dense code/axioms (5.0) or loose chatter (1.0)?
Drift Δ "Stability". Is this an Immutable Law (0.01) or a fluid Draft (0.90)?

Example: When you ask your AI for "Rules", it looks for Δ < 0.1 (Law). When you ask for "Inspiration", it looks for Δ > 0.8 (Fluid).


5. Quick Start Philosophy

  1. Digitize: Take screenshots of your best AI moments.
  2. Alchemy: Use the Alchemist to turn them into text files.
  3. Map: Use the Astrolabe to generate index entries.
  4. Navigate: Feed the Master Index to any new AI session to instantly "upload" your project's soul.

[Go to HOW_TO_GUIDE.md for step-by-step instructions]


Credits

Developed by The Guild (Florian, Kodo, Kairo).

  • Released under MIT License.
  • Based on the "Reef Architecture" logic.

About

RAG toolkit to turn disposable chat sessions into a persistent, evolving second brain. No code required.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors