1. Getting Started

1.1 What you need

Node.js (see the README for the supported semver range) and npm
An internet connection (first run only; needed to fetch the npm package and the browser used for PDF rendering)

Python is not required.

On the first run, pfpdf automatically downloads Chromium for PDF rendering. The download may be several hundred megabytes. Subsequent runs use the cached browser. On Linux, Chromium also requires certain system libraries (see Chapter 08).

By default, a five-minute timeout covers the entire process, from the initial browser download through PDF completion. If a slow connection or large document needs more time, increase --render-timeout-ms as described in Chapter 07.

1.2 Your first PDF

Create hello.md.

---
title: My First pfpdf
---

# Hello

This is my first PDF made with **pfpdf**.

Convert it.

npx @pfnet-research/pfpdf@latest --input hello.md --output hello.pdf

After the command succeeds, hello.pdf contains a cover with the title and a body with the heading and paragraph.

1.3 Pinning the version

In CI and long-lived runbooks, pin a specific version instead of using latest.

npx --yes @pfnet-research/pfpdf@0.1.0 --input hello.md --output hello.pdf

1.4 Converting a whole directory

When you pass a directory to --input, pfpdf combines the *.md files directly inside it into a single PDF, ordered by filename.

docs/
  00_title.md
  01_intro.md
  02_details.md

npx @pfnet-research/pfpdf@latest --input docs --output docs.pdf

Files are ordered by the byte sequence of their filenames. We recommend adding numeric prefixes such as 00_ and 01_.

1.5 When things go wrong

Use --doctor to diagnose your environment.

npx @pfnet-research/pfpdf@latest --doctor

The command checks Node.js, browser availability, fonts, and other parts of the environment, then reports any problems with suggested fixes. See Chapter 08 for details.