A system for creating academic content that can be rendered in multiple formats (slides, notes, papers) from a single source. LaMD provides a rich set of macros and contexts for handling different content types and output formats.

Full documentation is available at:

• https://inverseprobability.com/lamd

Key sections:

• Getting Started
• Context Reference
• Usage Guides

The system relies on the generic preprocessor, gpp (https://github.com/logological/gpp).

On Linux:

apt-get install gpp

On macOS:

brew install gpp

# Install lamd using pip
pip install lamd

# Or install from source using Poetry
git clone https://github.com/lawrennd/lamd.git
cd lamd
poetry install

To populate minimal agent hints (Cursor rules + AGENTS.md/CLAUDE.md blocks) in a LaMD-using repository (e.g. talks/, execed/), run this from that repo root. The installer writes files into the current directory (it doesn’t require any LaMD-specific files to already exist in the repo).

bash -c "$(curl -fsSL https://raw.githubusercontent.com/lawrennd/lamd/main/scripts/install-minimal.sh)"

To also create a local venv and install lamd (so maketalk is available without system Python), add:

bash -c "$(curl -fsSL https://raw.githubusercontent.com/lawrennd/lamd/main/scripts/install-minimal.sh)" -- --with-venv

LaMD requires:

• Python 3.11 or higher
• gpp preprocessor
• Pandoc (for document conversion)
• Additional Python packages (installed automatically):
  • lynguine
  • python-frontmatter
  • pandas
  • python-liquid

Create a _lamd.yml in your project root to configure pandoc flags and other settings:

LaMD uses a modular configuration system based on makefiles and markdown frontmatter:

Configuration is organized into flag files for different document types:

• make-cv-flags.mk: Configuration for CV documents
• make-talk-flags.mk: Configuration for talk documents
• etc.

These flag files:

1. Extract configuration from markdown frontmatter using mdfield
2. Define document-specific settings and paths
3. Are included by the make system as needed

Configuration can come from:

1. Markdown frontmatter (primary source)
2. Flag files (document-type specific)
3. Environment variables (for system-wide settings)

For a talk document:

1. The markdown file's frontmatter defines basic settings along with _lamd.yml that contains the defaults.
2. make-talk-flags.mk extracts these settings using mdfield
3. The make system includes these flags
4. The Python code (mdpp.py) respects these settings

LaMD provides several utility scripts:

• maketalk: Converts talk files from markdown to other formats
• makecv: Converts CVs from markdown to other formats
• flags: Extracts pandoc flags from _config.yml
• mdfield: Extracts fields from markdown headers
• dependencies: Lists dependencies in markdown files
• mdpeople: Generates people macros from YAML definitions
• mdlist: Generates lists from YAML definitions

The mdpeople script provides a powerful way to manage information about people in your documents. It generates macros for displaying profile images and information consistently.

Usage:

# Generate people macros from YAML file
mdpeople generate -i people.yml -o talk-people.gpp

# List all defined people
mdpeople list -i people.yml

# Check for missing images or invalid URLs
mdpeople verify -i people.yml

Example people.yml:

- given: Peter
  family: Piper
  image: people/peter-piper.png
  url: https://pepper-pickers.com/
  title: Neil Lawrence

- given: Jack
  family: Spratt
  image: people/jacks.png
  url: https://lowfatyoghurts.com/jackspratt
  title: Jack Spratt

Using generated macros:

\include{talk-people.gpp}

\section{Team}

Our team includes \neillawrencePicture{15%} and \carlhenrikekPicture{15%}

• Multiple output formats (HTML, PDF, PPTX)
• Context-aware content rendering
• Rich macro system for content reuse
• Media handling (images, videos, diagrams)
• Bibliography management
• People and team pages
• Exercise and assignment support
• Code testing and execution support (--code test)

Create a new talk:

# Create talk file
echo "---
title: My Talk
author: Your Name
---

\section{Introduction}

\notes{Detailed notes here}

\slides{
* Bullet points
* For slides
}" > talk.md

# Build different formats
maketalk talk.md --format slides
maketalk talk.md --format notes

LaMD supports code extraction and testing through the --code test option:

# Extract all code blocks for testing/execution
mdpp input.md --code test --macros-path macros/

# Use \testcode macro for test-specific code
\testcode{
import numpy as np
result = np.mean([1, 2, 3, 4, 5])
assert result == 3.0
print("Test passed!")
}

The --code test option enables all code inclusion flags (-DCODE=1, -DPLOTCODE=1, -DHELPERCODE=1, -DDISPLAYCODE=1, -DMAGICCODE=1, -DTESTCODE=1) to ensure all code blocks are available for extraction and execution.

• Source Code
• Issue Tracker

• Snippets: Repository of reusable content
• Talks: Example talks using LaMD