Campaign Archive Format

Technical reference for Mimir’s campaign export/import archive format.

Overview

Campaign archives are gzip-compressed tar files with the extension .mimir-campaign.tar.gz. An archive contains a complete snapshot of one campaign — metadata, modules, documents, characters, maps, tokens, homebrew content, and binary assets. Archives are self-contained and portable between machines and Mimir installations.

Property	Value
Format identifier	`mimir-campaign`
Format version	`2.0`
File extension	`.mimir-campaign.tar.gz`
Export filename	`<campaign-name-slug>.mimir-campaign.tar.gz` (lowercased, non-alphanumeric characters replaced with hyphens; consecutive separators collapsed to a single hyphen; leading and trailing hyphens stripped)

Archive Structure

<slug>.mimir-campaign.tar.gz
├── manifest.json                    # Archive metadata
├── data.json                        # All campaign data
└── assets/
    └── {asset-id}/{filename}        # One file per campaign asset

The archive contains exactly two JSON files plus one entry per asset. Asset files are stored under their original asset ID and filename; binary content is not hashed or deduplicated.

Manifest

manifest.json contains archive metadata:

Field	Type	Description
`version`	string	Archive format version (`"2.0"`)
`format`	string	Format identifier (`"mimir-campaign"`)
`created_at`	string	ISO 8601 export timestamp (UTC)
`campaign_name`	string	Campaign name at export time
`mimir_version`	string	Mimir version that created the archive
`counts`	object	Entity counts (see below)
`catalog_references`	array	Catalog items referenced by campaign documents

Counts

Field	Type
`modules`	number
`documents`	number
`characters`	number
`maps`	number
`tokens`	number
`module_monsters`	number
`module_npcs`	number
`assets`	number
`homebrew_items`	number
`homebrew_monsters`	number
`homebrew_spells`	number

Catalog References

Each entry identifies a catalog item that campaign documents reference but the archive does not contain:

Field	Type	Description
`type`	string	Reference type: `monster`, `spell`, `item`, `creature`, `condition`, `feat`, `race`, `class`, or `background`
`name`	string	Item name
`source`	string	Source book code

References are extracted at export time by scanning document content for {@type name|source} tags.

Campaign Data

data.json contains all campaign entities in a single JSON object:

Field	Type	Description
`campaign`	object	Campaign record
`sources`	string[]	Enabled source book codes
`modules`	array	Module records
`documents`	array	Document records (full text content)
`characters`	array	Characters with related data (see below)
`maps`	array	Maps with related data (see below)
`tokens`	array	Token placements across all maps
`module_monsters`	array	Monster references attached to modules
`module_npcs`	array	NPCs attached to modules
`assets`	array	Asset records (filename, MIME type, blob path)
`homebrew_items`	array	Campaign homebrew items
`homebrew_monsters`	array	Campaign homebrew monsters
`homebrew_spells`	array	Campaign homebrew spells

Character Entries

Each character entry contains the base character fields (flattened) plus:

Field	Contents
`classes`	Class entries (levels, subclasses, starting-class flag)
`inventory`	Inventory items
`spells`	Known spells
`proficiencies`	Proficiency entries
`features`	Class features
`feats`	Feats

Map Entries

Each map entry contains the base map fields (flattened) plus:

Field	Contents
`pois`	Points of interest
`traps`	Trap placements
`light_sources`	Light source definitions
`fog_areas`	Fog of war revealed areas

Asset Files

Binary files (map images, token art) are stored in the archive at assets/{asset-id}/{filename}, where {asset-id} is the asset’s UUID at export time and {filename} is its original filename.

Import Behavior

When an archive is imported:

The archive is extracted and data.json is parsed; import fails if data.json is missing or unparseable. The manifest’s format field is not checked during import — that validation happens only in archive preview (see Preview)
New UUIDs are generated for all entities, so imports never conflict with existing data
Internal references (module, document, character, map, asset, and monster/NPC IDs) are remapped to the new UUIDs
Asset files are copied into the importing installation’s assets directory under campaigns/{new-campaign-id}/{filename}
Campaign sources are restored; they take effect only if the corresponding catalog data exists in the importing installation
If the campaign name already exists, a numeric suffix is appended (e.g., Name (2)), unless a name override is supplied

Importing the same archive twice creates two independent copies of the campaign.

Preview

An archive preview reads manifest.json without extracting or importing the rest of the archive. Preview is where the manifest’s format field is validated: archives whose format is not "mimir-campaign" are rejected. The preview contains:

Field	Description
`campaign_name`	From manifest
`counts`	Entity counts from manifest
`catalog_references`	Catalog items the campaign references
`mimir_version`	Mimir version that created the archive
`created_at`	Export timestamp
`archive_version`	Archive format version