md² (The mdd Engine)

md² (pronounced md-squared) stands for mdBook deepDive. Much like vim is a recursion of vi, md² is a recursive evolution of the mdbook ecosystem, transformed into an opinionated platform for the Personal Knowledge Blockchain.

In the age of LLMs, the cost of generating information has collapsed to zero, but the cost of maintaining Signal has skyrocketed. Most research dies in transient chat windows.

md² is the engine for those who refuse to let their inquiries evaporate. It provides a formal framework to organize research into an immutable ledger of discovery:

• Block 0 (The Genesis Block): The hardcoded rules of your universe. It contains the Six Foundational Pillars (e.g., Bitcoin, Intelligence, Physics) that anchor all subsequent research.
• Knowledge Blocks (Block 1, 2, ...): Chronological epochs of discovery. Each block is limited by the Law of 21; once the 21st episodic transaction is mined, the block is locked into the permanent chain.
• The Mempool: The high-entropy zone for unconfirmed drafts and raw signal.

Research in md² follows a "Bottom-Up Mining, Top-Down Reading" flow. As insights are discovered within individual episodes, they are "bumped up" to the Block 0 pillars. Over time, the Genesis Block becomes a compressed, high-fidelity synthesis of your entire knowledge chain.

md² transforms raw LLM intelligence into a "Clean Internet" publication that spans three densities:

1. The Slice (Visual): Cinematic video infographics (short-form) used as "Visual Covers."
2. The Cake (Audio): High-level audio overviews for podcast distribution (Spotify, Apple, Fountain).
3. The Dough (Text): Deep, KaTeX-hardened mathematical research papers anchored on mdbook.

To prevent "LLM Pruning" during export, use the Master Packaging Prompt to force the LLM to generate a self-extracting Python payload. This ensures 100% structural integrity and zero-loss transmission of complex math and citations.

Run the md-publish binary to "mine" the payload into the chain. The engine automatically:

• Hardens KaTeX: Fixes math delimiters and escapes currency for mdbook-katex compatibility.
• Re-indexes Citations: Manages sequential footnotes and duplicate references.
• Syncs Sidebar: Enforces the 3-layered 4-space hierarchy in SUMMARY.md.
• Injects Multi-modal UI: Adds the GitHub download CTA, lightning widgets, and video carousels.

• Rust & Cargo: (v1.88.0+)
• mdbook: (v0.5.3+)
• mdbook-katex: (v0.10.0-alpha+)
• Python 3: For payload extraction.

# 1. Clone the mdd engine
git clone https://github.com/ashutoshmjain/mdIngest.git
cd mdIngest

# 2. Build the mining tool
cargo build --release

# 3. Add to Path
cp target/release/md-publish /usr/local/bin/

[preprocessor.ingest]
command = "md-publish"
downloads_path = "/path/to/your/downloads"
video_source = "/path/to/your/video/archive"
lightning_address = "yourname@primal.net"
title_word_limit = 5

Copy this into your LLM session to generate a lossless export:

Objective: Convert the exhaustive research conducted in this session into a self-extracting Python payload to ensure 100% structural integrity and zero-loss transmission.

1. Packaging Requirements:
* Full Fidelity: Package the exhaustive research in its entirety. Do not summarize or prune.
* Absolute KaTeX: Wrap all mathematical notation and symbols in Absolute KaTeX delimiters: $...$ for inline and $$...$$ for block displays.
* Hyperlinked Bibliography: Format every entry in the "Works Cited" section as a clickable Markdown link: [^N]: [Author, Title, Year](URL).

2. Technical Encoding:
Generate a Python Script that performs the following:
1. Assign the complete Markdown text to a variable named payload_text.
2. Gzip-compress and Base64-encode the payload_text.
3. Output the final Python script containing this encoded string and the logic to decode and write it to a file named final_research.md.

Licensed under CC0 1.0 Universal. Build the future.