Publisher websites save scientific manuscripts under cryptic file names like s123456.pdf. This is not helpful because one can't tell what's in the file without opening it. This workflow gives each downloaded paper a readable name built from its authors, journal, and year. It's for anyone who collects papers from journal sites and needs to find them again later.

1. Download the manuscript from the publisher's website (for example Nature).
2. Use AI to read the file and pick out the first author, the second author, the journal, and the year.
3. Rename the file to FirstAuthor_SecondAuthor_Journal_Year.pdf.

Result: each saved paper can be identified from its file name alone, without opening it.