Use a browser agent to find and save papers

A commenter uses a custom browser agent for a narrow set of research and data-entry jobs, including finding scientific papers and downloading bank CSVs.

Done with Custom agent· Intermediate· Jul 2026

What they did
They built their own browser agent and keep it on a tight leash. For paper research, it searches for relevant scientific articles and saves open-access PDFs. They also use the same setup for pulling bank CSVs and entering company transactions, with fixed Python steps handling part of the workflow and AI making the rest of the choices.
How it went
They describe it as useful for a limited set of tasks, but not something they let run on its own. No performance numbers or failure examples were given.
Tips
Keep the job narrowly defined and let code handle the repeatable parts. For anything that could cost money or change records, pause and ask before the agent sends, submits, or deletes anything.

Try it yourself with Claude in Chrome

Help me research [topic] using [Google Scholar/arXiv/PubMed]. Find the [number] most relevant recent papers, open each result, and download any open-access PDFs into a folder called [folder name]. Build a list with the title, authors, year, and a two-line summary for each paper. If a paper is paywalled or unclear, stop and ask me whether to skip it or try another source before downloading anything. Before you submit, send, delete, or otherwise make any irreversible change, pause and ask for my approval.

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

Get 3 agent uses like this every morning

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