Literature Search

Literature Search

Literature Search uses the OpenAlex database to find papers. It is available under Find papers.

Search

Enter one or more Keywords. They are combined into the OpenAlex query; a phrase can remain one keyword. Optional Qualities describe what makes a paper useful beyond matching the query.

When scoring is enabled, RxCubed combines keyword and quality signals using the configured 0.5/0.5 weighting. The result shows AI score, search rank, an AI summary, and scoring details. The AI summary combines two short sentences: what the paper is about and why it is relevant to the query and qualities.

A paper with no abstract in OpenAlex is not scored — a title alone isn't enough signal. It keeps its original OpenAlex search position, and scoring only re-ranks the papers that do have an abstract into the remaining slots.

OpenAlex searches are limited to 50 per user per day. A warning appears at 40 searches. If you need more, add your own OpenAlex API key in Settings.

Review and shortlist

Search results retain the paper title, authors, year, source, abstract when available, and source link. Extra OpenAlex metadata is under Details. DOI and other identifiers remain available there rather than crowding the main result.

Click Add to shortlist on each paper you want, and choose its file type (Literature Search always offers PDF). Nothing becomes project evidence until Save shortlist, at the bottom right of the Shortlist card.

Files

RxCubed attempts to fetch the PDF automatically when the shortlist is saved.

Open access is prioritised. OpenAlex often knows several copies of the same paper — a repository, PubMed Central, the publisher — and RxCubed prefers a trusted open-access copy over the publisher's own page, judged on where the file actually lives rather than on a metadata flag alone. If the first copy fails, whether because the link is dead or because the host refuses automated requests, it tries the other known copies before giving up.

A missing or paywalled PDF is still a valid evidence record. It appears under Missing Files, where file ↗ opens the source and Manual Upload attaches what you downloaded.

Asking the assistant about your results

The AI Assistant can see the results currently on your screen — each one's title, search rank, AI score, and AI summary — so you can ask it which to save rather than reading every card. If a question needs more than the summary for one paper, it can read that paper's full abstract itself. See AI Assistant.

Related views

All Files is the Literature Search slice of Results Hub and uses the same Category → Tool → Run → Evidence organisation. Use Results Hub for the complete project view. There is no separate export view on this page.