If your campaign already lives in Foundry VTT, you do not have to retype it. Foundry can write any journal entry, actor, item or scene out as a JSON file, and The Chronicler reads those files directly.
We never connect to your Foundry server and we never ask for a Foundry login. You export the files yourself and upload them — that is the whole arrangement.
Export from Foundry
Foundry's export is a right-click away. In your world:
- Open the Journal (or Actors, Items, Scenes) sidebar tab.
- Right-click a document → Export Data. Foundry saves a
.jsonfile such asfvtt-JournalEntry-the-dead-marshes.json. - Repeat for everything you want imported, or install one of the community bulk exporters (Easy Exports, Foundry Bulk Exporter, Adventure Importer/Exporter) and export a whole folder as one
.zip. The Chronicler reads the ZIP and the loose JSON files identically.
The filename does not matter. The Chronicler recognises Foundry documents by their shape, so renamed files still import.
Upload to The Chronicler
- Go to the campaign Overview.
- Open Bootstrap campaign.
- Drop the
.jsonfiles, or the export.zip, into the upload box and press Start bootstrap.
You land on the bootstrap review page while it processes.
What gets read
| Foundry document | What The Chronicler takes |
|---|---|
| Journal entry | The entry name, every page name, and each page's text. This is the prose the wiki is built from. |
| Actor | The name, the Foundry actor type, and the biography or notes field. An actor of type npc becomes an NPC candidate; type character becomes a player character candidate. |
| Item | The name and description — an item candidate. |
| Scene | The name only, as a location candidate. A scene is a map, not prose. |
Any other actor type — a vehicle, a group, a hazard, a type specific to your system — is not typed for you. Its text still goes through, and the AI may propose it as an entity of whatever type fits; we would rather read your prose than guess a label.
Biographies are read system-agnostically. A D&D 5e actor's biography, a Pathfinder 2e actor's public notes, and a bespoke system's description field all work; if nothing biography-shaped is there, the actor still imports with its name.
Anything in a ZIP that is not a readable document — images, maps, module code, audio — is skipped. A JSON file that is not a Foundry document is still read as text, so exports from other note tools usually work too.
Nothing is added until you approve it
Everything an import proposes arrives as a candidate on the bootstrap review page. You rename, retype, keep or dismiss each one, and only then does it become a wiki entry. Imported text is AI-owned prose: it is never written as canon, and it can never overwrite a summary you have approved or a name you have locked. See The GM is the authority on canon.
Every entity that comes from an import records where it came from — that it was a Foundry import, which file it was in, and the Foundry document's own id — so a wrong import is findable and removable rather than a mystery six months later.
Size limit
One bootstrap run reads up to 100,000 characters of text in total, and a single file inside a ZIP must be under 5 MB. A long-running Foundry world will be bigger than that, and the run will tell you so rather than importing a third of your world silently. Export in batches — your world book first, then your NPC folder — and run bootstrap once per batch.
What we deliberately do not do
There is no "connect my Foundry account" button, and there never will be: your world is self-hosted and your data is yours, so a file you export is both the safest and the most complete route. For the same reason we do not offer a D&D Beyond or Roll20 account importer — see Import a transcript or notes for the routes those platforms do sanction (a D&D Beyond PDF character sheet, a Roll20 chat archive you saved yourself).