Literature Radar

Self-hosted literature monitoring for scientists. It watches arXiv, PubMed, Crossref, bioRxiv, and medRxiv for your topics, has your own AI score every new paper against a plain-language profile of what you care about, and emails you a clean digest of the ones worth reading, each with a one line reason why it surfaced. One PHP app, one SQLite file.

How it works

1. You describe your research interests in plain language and add topic search queries.
2. A cron job polls the sources on your schedule and collects new papers.
3. Your own language model scores each paper 0 to 10 against your interest profile, with a short reason.
4. On your chosen schedule you get one email with the best papers, best first. Nothing repeats, and everything stays browsable in the admin area with score and source filters.

Bring your own model

Ranking runs through whichever of these you have, checked in this order:

1. claude CLI, so a Claude subscription covers ranking with no per-token cost
2. codex CLI, same idea for a ChatGPT subscription
3. Anthropic API key
4. Any OpenAI compatible API (OpenAI, OpenRouter, Ollama, LM Studio)

Setup

cp config.sample.php config.php

Fill in an admin password, your digest email, and SMTP details, then either serve public/ with PHP or use Docker:

docker compose up -d

Add the cron job:

0 */2 * * *  php /path/to/Literature-Radar/cli/cron.php

Each run polls when a poll is due, ranks anything new, and sends the digest when it is due. php cli/cron.php poll, rank, or digest runs a single step by hand.

Sources

• arXiv for physics, math, CS, and quantitative biology
• PubMed for biomedical and life sciences (add a free NCBI key in config for a higher rate limit)
• Crossref for the rest of the journal world
• bioRxiv and medRxiv preprints, matched locally against your topics

Polling is polite by design: requests identify themselves with your contact email, arXiv queries are spaced out, and the poll window and frequency are yours to set.

The admin area

public/admin.php gives you the ranked paper list with score and source filters, topic management, the interest profile editor, a live digest preview, status counts, and one-click poll, rank, and send buttons.

Author

Built by Paul Crinigan, AI Apps API. Full documentation lives on the Literature Radar page, part of the free AI science tools collection at learnhowtoscience.com.

MIT licensed.