In COBOL, a copybook is the file that describes a record: which bytes hold the account number, and how many decimal places sit inside the balance. Mainframe data files are unreadable without it. Any bank or insurer moving data off the mainframe starts by digging out its copybooks.
Copybook.org takes that working term, which is also an ordinary English word, and puts it on the extension serious infrastructure projects favor. SQLite lives at sqlite.org. Nginx lives at nginx.org. A tool that becomes the standard way to read mainframe data belongs in that company.
The problem behind the name
Mainframe files don’t open like a CSV. The text is in EBCDIC, an IBM character set, and the numbers are often stored as packed decimals: two digits to a byte, with the sign tucked into the last half-byte. Get the copybook wrong by one field and every record after it shifts into nonsense.
The tool this name wants is a single binary plus a browser demo. Drop in a copybook and a data file, and get JSON, CSV, Parquet or SQL back, packed decimals and all. It’s the first job in any migration, and right now teams do it with scattered scripts and vendor how-to pages.
Modernization budgets are big
COBOL still runs core systems at banks and insurers, and moving off it is a multi-year, big-budget project wherever it happens. Cloud providers run whole programs to win those migrations, and consultancies build practices around them. A clean, trusted tool for the first step, under a name every mainframe engineer recognizes, has an audience with real money behind it.
It suits a modernization consultancy, a cloud migration vendor, a data-integration company or an open-source project that wants a .org to grow into.
Every migration opens the copybooks first.