Making a Folder of Recipe PDFs Actually Useful

My wife has been following a meal plan service that releases new PDFs every few weeks. (shout-out to the Health With Bec programme, she's got some great recipes!) Over time we had accumulated a large folder of these: dozens of plans, hundreds of recipes, each buried somewhere inside a multi-page document. Finding a specific recipe meant either remembering which plan it was in or scrolling through PDFs one by one. Neither was fast.

What my wife really needed to make it usable and easy was a single view across the whole collection: search by name or ingredient, filter by meal type, and jump straight to the right PDF and page. That felt like a weekend project. It turned out to be a single conversation.

What the Artifact Does

The result is a recipe library that runs entirely in the browser. Every recipe from the collection is in there as a card: name, meal type, calories, macros, which PDF it came from, and a direct link to that document. A search bar filters across recipe names and sources. Tabs along the top filter by meal type. The data is embedded in the file itself, so there is no backend, no login, no dependency on anything external staying online.

The harder part was getting the data out of the PDFs in the first place. Claude handled that too, reading through each document and pulling out the structured information needed to populate the library. The whole thing, from raw PDFs to working artifact, took one conversation.

The Pattern Worth Noting

This use case sits in a gap that often gets overlooked. The data already existed. The PDFs were already there, organised, complete. The problem was purely about access: getting from "I want that salmon recipe" to the right page in the right document without three minutes of hunting. A Claude artifact is a good fit for exactly this kind of problem. The effort is in describing what you want; the output is a working tool you can use immediately and share with anyone. Right, I'm off to find a snack that uses bananas.