How it works
Paste any messy reference list - copied from a PDF, a blog post, an email - and the parser pulls out author, title, publisher, year and URL from each line. You get a per-source audit plus three clean outputs: AI-friendly markdown, schema.org JSON-LD for structured data, and CSV for spreadsheets and knowledge bases.
Why normalised sources win
- One shape, all parsers. Retrieval layers score predictable shapes higher - same fields in the same order every time.
- No silent gaps. The audit flags missing years, authors and URLs before they become silent retrieval failures.
- Schema-ready output. The JSON-LD format drops straight into an ItemList in your page head for machine-readable signal.
- Zero manual cleanup. Paste 50 lines, export cleanly - what used to take an hour now takes a few seconds.
Pair with
Feed the clean output into the Reference Generator to format for APA or MLA, the Citation Schema Generator for structured markup, or the Evidence Block Generator to wrap individual sources around live claims. Context: how assistants choose citations.