Skip to content
Get AI Business OS

Collections and Tags

PRIMA Scholar

Collections and tags are two ways to organise the records in Your Research Library. A collection is a named group, such as one per project or thesis chapter. Tags are short labels you attach to individual records, such as a method or a theme. Each record can belong to any number of collections and carry any number of tags.

Ask Claude to create a collection and give it a name, with an optional description of its purpose. Claude uses the library_create_collection tool. Each collection name must be unique, so a second collection with a name already in use fails with an error.

Create a collection called Spring Workshop Reading for the P012 workshop sources.

Ask Claude to add a record by its document ID and the collection name. Claude uses the library_add_to_collection tool, which needs the collection to exist already. The name must match exactly, including capital letters, or the tool reports that the collection was not found. If the record is already in that collection, Claude reports that nothing changed.

/library import offers to add each newly imported file to a collection. It lists the collections you have and offers to create a new one when none fits.

/library collections lists every collection in name order, with names that start with a capital letter before names that start with a lower-case letter, together with its description and the number of records it holds. Claude then offers to browse one collection or create a new one.

Browsing a collection uses the library_list tool with the collection name. The list shows the newest added records first, 20 at a time, and you can ask for the next page.

Ask Claude to tag one or more records by document ID. Claude uses the library_tag tool once for each record and creates the tag the first time you use it. Removing a tag works the same way and takes the tag off that record only.

Tag documents 12, 15 and 23 as motivation.

Tag names must match exactly, including capital letters. Burnout and burnout are two different tags, so choose one form and keep to it.

/library tags shows your most used tags with the number of records carrying each one. The list holds the top 20 tags, and Claude tells you when the list reaches 20. To list the records with one tag, ask Claude to filter the library by that tag.

/library stats shows the same top tags alongside totals by file type and by year.

PRIMA Scholar 2.2.1 has no tool for these changes, so Claude cannot make them for you.

  • Remove a record from a collection.
  • Rename or delete a collection.
  • Rename a tag, or delete a tag from every record at once.
  • List every tag when you have more than 20 in use.

Deleting a record does take it out of every collection and removes its tags, as part of the deletion.

  • Tag consistently: use lowercase, hyphenated tags such as self-determination-theory or qualitative-methods, and avoid synonyms.
  • Use collections for projects: create one collection for each writing project, thesis chapter or research question.
  • Add a note on import: a one-line note on why a paper matters saves time when you return to it months later.
  • Review now and then: /library stats and /library collections give you counts to check the library against.

The Literature Watch uses these features too. By default it tags every paper it imports literature-watch and adds it to a collection called Literature Watch, which it creates when needed.