explaingit

marcj/papernews

Analysis updated 2026-05-18

107PythonAudience · generalComplexity · 2/5Setup · moderate

TLDR

A personal tool that turns RSS feeds and Hacker News into a single daily PDF styled like a quiet academic paper, using Claude to summarize and rewrite articles.

Mindmap

mindmap
  root((repo))
    What it does
      Fetches articles
      Summarizes with Claude
      Rewrites and translates
      Renders PDF
    Tech stack
      Python
      Docker
      LaTeX
    Use cases
      Read news offline
      Send to e-reader
      Customize the template
    Audience
      Individuals
    Setup
      Docker
      Anthropic API key

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

What do people build with it?

USE CASE 1

Generate one daily PDF combining RSS feeds and Hacker News with no ads or clutter.

USE CASE 2

Read full article text offline instead of switching between many websites.

USE CASE 3

Send the daily PDF to an e-ink device like the reMarkable tablet.

USE CASE 4

Customize the LaTeX template or rewrite prompt to change the look and voice.

What is it built with?

PythonDockerLaTeXClaude API

How does it compare?

marcj/papernewspython/tzdatatencentyouturesearch/t2i-l2p
Stars107107107
LanguagePythonPythonPython
Last pushed2026-07-01
MaintenanceActive
Setup difficultymoderateeasyhard
Complexity2/51/54/5
Audiencegeneraldeveloperresearcher

Figures from each repo's GitHub metadata at analysis time.

How do you get it running?

Difficulty · moderate Time to first run · 30min

Runs in Docker and requires an Anthropic API key for the summarize and rewrite steps.

In plain English

papernews is a personal news reading tool that converts RSS feeds and Hacker News into a single PDF, formatted to look like a quiet academic paper or old magazine rather than a web page. The author built it because they disliked switching between multiple sites with different layouts, advertisements, and visual clutter. The result is one PDF per day with consistent typography, no color, and full article text rather than summaries, readable entirely offline. The tool works in four stages. First it fetches the full article text from each source you configure. Second it uses Claude (an AI model by Anthropic) to produce a short two-sentence summary of each article. Third it asks Claude to rewrite the full article body into clean, consistently styled prose and to translate any non-English content into English if you want that. Fourth it renders everything through a LaTeX typesetting system into a PDF with a cover page, table of contents, two-column article layout, and proper formatting for math and code. Setup runs in Docker. You clone the repository, paste your Anthropic API key into a configuration file, edit a sources file to list which RSS feeds and how many articles from each, then run the container. A web server at localhost port 8000 serves the current PDF. The first build takes a minute or two, after that results are cached and a background process refreshes the content every four hours. The PDF is designed to be sent to an e-ink device like the reMarkable tablet. The README describes a few ways to do this: manually downloading and uploading through the device's web interface, using a third-party command-line tool on a cron schedule, or emailing the file to an email-to-device bridge. You can customize the output by editing the LaTeX template to change page size, fonts, or layout, and by editing the rewrite prompt to change the editorial voice or disable the English translation step. The project is described as a hobby tool that works but may have rough edges.

Copy-paste prompts

Prompt 1
Explain the four stages this tool uses to turn RSS feeds into a PDF.
Prompt 2
Help me set this project up in Docker with my Anthropic API key.
Prompt 3
How do I configure which RSS feeds and how many articles to include?
Prompt 4
Show me how to change the rewrite prompt to alter the editorial voice.

Frequently asked questions

What is papernews?

A personal tool that turns RSS feeds and Hacker News into a single daily PDF styled like a quiet academic paper, using Claude to summarize and rewrite articles.

What language is papernews written in?

Mainly Python. The stack also includes Python, Docker, LaTeX.

How hard is papernews to set up?

Setup difficulty is rated moderate, with roughly 30min to a first successful run.

Who is papernews for?

Mainly general.

Open on GitHub → Explain another repo

This repo across BitVibe Labs

Verify against the repo before relying on details.