Fantasy tabletop realm wiki
Build living campaign realms with notes, lore, maps, and AI.
DMRealm is a friendly pen-and-paper style workspace for game masters and players: a public wiki, a private prep desk, a visual atlas, character and shop studios, and AI tools for turning loose ideas into organized campaign material.
A wiki that knows the table has secrets.
Public pages can be shared and indexed. Private pages stay private. DMs can keep drafts, clues, reveals, stat notes, and campaign-only material away from players until the right moment.
Realm wikis
Publish a public realm wiki, keep private campaign notes hidden, and give players access only to what they should know.
AI-assisted prep
Draft NPCs, encounters, lore, locations, maps, handouts, and recap notes while keeping generated material private until reviewed.
Connected context
Link pages, factions, characters, quests, sessions, locations, assets, and secrets into a relationship graph.
Player notes
Let players keep personal theories, character memories, clue boards, and shared party notes without exposing DM material.
Creation studios
Build from pages, maps, characters, shops, or whatever sparks the next session.
A realm can grow from a wiki page, a sketched border, a placed city marker, a merchant's limited shelf, or a character description. DMRealm keeps the structured data connected so future AI drafts have useful context.
Atlas canvas
Create world maps, city maps, dungeon plans, combat scenes, cosmology diagrams, pantheon boards, faction webs, and clue maps from one visual system.
Character studio
Generate profiles, stat block drafts, images, tokens, equipment, relationships, and multilingual behavior notes for NPCs, villains, merchants, and creatures.
Shop studio
Build player-facing shops with limited stock, prices, restock rules, hidden inventory, random supplies, specific items, and merchant links.
Beyond MVP
Start as a campaign notebook. Grow into a realm operating system.
Designed for connected context
Every page, canvas feature, character, shop item, image, session, quest, and clue can become context for future generation, but only if the requesting user has permission to see it.
Commercially cautious
Original fantasy content, no official publisher branding, and clear review points before AI material becomes public.
Mock-first AI
Local development works without paid AI APIs, while provider adapters keep vendors replaceable.
Built for CI/CD
Typed packages, tests, production builds, and Cloud Run-oriented services are ready for pipeline setup.