articrab

articrab is a CLI tool to fetch articles and structure them as Markdown.

Install

Requires Node.js >= 22.

Global install:

npm install -g articrab

Then run:

articrab --help

Usage

Important: set the Jina token first.

articrab set-jina-token jina_xxx

Fetch an article:

articrab <url> --output ./content

Flow

URL -> fetch -> extract -> scoring -> html2md -> write output
             |                 |   |
             |                 v   | yes
             |                good enough?
             |                     | no
             |                     v
             <---------------- jina fetch

Examples

Example output is available in:

• 20260331-134024-912-harness-engineering-leveraging-codex-in-an-agent-first-world

Test

npm test