A Folder Says Where Something Lives. A Tag Says What You Still Need to Ask.

At forty people, you remember everyone. You know which Maria is the one from the Kraków baptism, which branch you have barely started, and which great-uncle has two conflicting death dates you meant to sort out.

At four hundred people, you do not. The archive is still all there, and search still finds anyone by name. What has gone is the layer on top: which people belong together for the piece of work you are doing this month, and which ones you flagged in your head as unfinished and have now forgotten about.

Researchers fill that gap in different ways, often with a spreadsheet, a notebook, or a surname prefix. Folders and tags are our version of it, and they are live today.

The short answer. KleioBase, an AI research companion that turns scattered historical records into one connected family archive, now lets you sort people and records into folders and label them with coloured tags, on every plan including the free one. Use them for two different jobs. A folder says where something lives: one folder per person or record, nested up to 10 levels deep, for a branch, a place or a project. A tag says what you still need to ask: as many as you like per item, shared between people and records, for a research status such as "To research" or "DNA match". The design rule that matters most is that a folder never cuts a connection. A record filed in one folder is still linked to every person it names, wherever they are filed, and duplicate matching still looks across the whole workspace.

Every filing system eventually hits the same wall

This problem is older than KleioBase. Our guide to organising family photos and documents digitally puts it bluntly: a wedding photo of your grandparents belongs in both their folders, and a letter mentioning six cousins belongs in seven. Folders enforce a single hierarchy, and family documents do not have one.

That article's answer was metadata: label each item with what it is about, so you can find it along any axis. That is still right. The trouble is that the labels you most need are not the ones written on the document. They are about your research. "Unproven." "Ask cousin Ann." "Check against the 1901 census." A folder of files has nowhere obvious to hold those, so they tend to end up in a separate notebook that slowly stops matching the files.

A folder answers where, a tag answers why

The two tools look similar and are not, and using them for the same job is how an organising system turns into clutter.

A folder is a place. Each person or record sits in at most one, which is exactly what makes a folder useful for answering "where is this?". Good folders follow the shape of your research: a folder per branch (the Cork side, the Quebec side), per place (everyone from one Ohio county), or per project (the line you are writing up for a family reunion). People and records have separate folder trees, which suits the fact that the way you group records, by archive or by record type, is rarely the way you group people.

A tag is a question. An item can carry any number of them, and the same tags are shared across people and records. So tags are where research status belongs: "To research", "Conflict", "DNA match", "Waiting on archive". Filter by one tag to see everything waiting on that question, or by several to see the overlap, choosing Any for items with at least one of the tags and All for items with every one.

The test for which to use is simple. If an item can only be in one place at a time, it is a folder. If it can be in several states at once, it is a tag.

In a connected archive, a folder cannot break a link

This is the part that makes folders in KleioBase different from folders on your hard drive, and it is deliberate.

On a computer, filing a scanned baptism in the "Cork" folder hides it from the "Kraków" folder. In KleioBase, a record is not just a file. It is linked to every person named in it, and those links do not care about folders. A marriage record filed under the Quebec branch still appears on the profile of the witness you filed under a different branch, and still sits on that witness's timeline. Moving a record never moves or removes the people it names.

The same is true of matching. Duplicate detection compares people across the whole workspace, not folder by folder, because the most valuable duplicate is precisely the one you filed in two different places without realising they were the same man.

That is also the line between folders and workspaces. We argued in why two families should never share one archive that the only thing that keeps an unrelated family out of your matches is a boundary drawn below the level where connecting happens. A workspace is that boundary. A folder deliberately is not. Folders are for organising one family's research; workspaces are for keeping two families apart.

What we built

Folders and tags are on the People and Records pages of the Knowledge Base, on every plan, including free Explorer. If you never create one, nothing changes.

Browsing. Click a folder to open it, and follow the path at the top back up. Searching at the top level searches everything, including what is inside folders; searching inside a folder searches that folder and everything below it.

Filing. Select people or records with their checkboxes and choose Move to folder, or drag them onto a folder or onto the path at the top. A single bulk action handles up to 500 items. Each person's page and each record's panel also has its own folder and tag controls.

At upload. Pick a folder from Save to before you upload, and new records land there. While you review a record before confirming it, each person in it has folder and tag controls beside their name, applied when you click Confirm, so the new people arrive already filed.

Tags. Type a new name to create a tag and pick its colour. Manage tags lets you rename, recolour or delete one. Deleting a tag only removes the label, and deleting a folder never deletes what is in it: its contents move up one level.

The Research Companion knows about them. You can ask the Research Companion something like "who is tagged To research in my Kraków folder", and it can create folders, move people and records, and create or apply tags when you ask, each one shown to you for confirmation before it happens.

Your folders and tags are included when you download a copy of your data. The Folders and tags guide covers every control.

Now the honest limits. Three of them.

They do not travel in a GEDCOM file. A GEDCOM export carries your people, families, events and sources, but not your folders or tags. If you move your tree to another program, the organisation stays behind.

The Companion cannot delete them. It can create and apply, but deleting a folder or a tag is done by you in the Knowledge Base.

There are ceilings. A workspace can hold up to 500 folders and 200 tags, and folders nest at most 10 levels deep.

The practice, whatever you use

The feature is ours. The habit is not, and it works in a spreadsheet as well as anywhere.

Tag the question, not the subject. The archive already knows who a record is about. What it cannot know is what you were going to do next. A tag such as "To research" or "Conflict" is a note to your future self that cannot drift away from the thing it is about.

Keep folders shallow. Ten levels are available. Two or three are usually enough. A folder you have to remember the path to is a folder you will stop using.

Tag the neighbours. When you work a brick wall through the people around your ancestor, give the witnesses, sponsors and neighbours one shared tag. That cluster is the raw material of FAN club research, and a tag keeps it together without filing anyone away from their own family.

Clear tags when the question is answered. A tag that never comes off is not a status any more. It is decoration.

A family archive is built by the people who wrote the records, and organised by the person reading them. The records do not need folders. You do, because you are the one who has to come back to them next month and remember where you were.