Review a paper folder with Claude Code

Use Claude Code to pull reference data and study summaries from a batch of PDFs, then rank them against a topic.

Done with Claude Code· Intermediate

What they did
The researcher set up two Claude Code slash commands and stored the topic in CLAUDE.md. /lit-map scans every PDF in a papers/ folder, extracts each paper’s aim, approach, findings, and keywords, and writes a ranked relevance file called literature_map.md. /citation-network reads the bibliographies, finds sources that show up in several papers, and orders them by frequency.
How it went
The author says /lit-map can turn a 50-paper review into 30 minutes, and /citation-network can cut a 10-paper review to 15 minutes. Those figures are claimed comparisons, not measured benchmarks, and the researcher still needs to check the AI’s outputs.
Tips
Scanned PDFs and image-only files are more likely to produce errors, so verify the results by hand. If the folder contains unpublished or copyrighted work, check the platform’s rules before uploading it.

Try it yourself with Claude Code

In Claude Code, go through every PDF in [path to my papers folder]. For each document, extract the bibliography plus the paper’s goal, method, main conclusion, and keywords. Write the bibliographies to bibliographies.md. Then create relevance_list.md that sorts the papers by how closely they match [my research theme], with one short reason for each item. Also identify the sources that appear in multiple papers and tell me the top 10 most frequent ones. Do not edit or delete any source files. Before you save or replace any output file, stop and ask for my approval.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on note.com ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.