You have a world. It lives in documents: a story bible, character notes, a wiki, maybe a folder of PDFs and a few years of scattered ideas. Your game engine cannot use any of it. Engines run on structured data, tables of entities with typed fields and explicit relationships, and prose is none of those things. The gap between "the world is written" and "the world is in the game" is where a lot of narrative projects stall.
This guide is the practical path across that gap: how to turn written lore into game-ready data, step by step. It works by hand, and it works faster with tooling. We will do it the manual way first, then show where a tool like Architect fits.
The core problem: prose is not data
Take a sentence like "Kaelen, once the queen's spymaster, now runs the smugglers of the Lower Docks and hates the man who exiled him." It is rich for a reader and useless to an engine. To your game it needs to become an entity (Kaelen) of type Character, with fields (role: smuggler; former role: spymaster; location: Lower Docks; status: exiled) and relationships (hates: the man who exiled him; controls: Lower Docks smugglers). Same information, different shape. Turning the first form into the second, across a whole world, is the job.
Step 1: Inventory your lore
Gather everything in one place: story bible, character sheets, location notes, faction histories, timelines, loose documents. You are not editing yet, just collecting. The goal is to know the full surface area before you structure it.
Step 2: Define your entity types and fields
Decide what kinds of things your world contains: characters, locations, factions, items, events, quests. For each type, decide the fields that matter to your game. A character might need role, allegiance, status, and location; an item might need owner, rarity, and effect. Keep it to what the game will actually use. Extra fields are maintenance you will not thank yourself for.
Step 3: Extract the entities
Go through the lore and pull out every instance of each type. Kaelen becomes a Character record, the Lower Docks becomes a Location, the smugglers become a Faction. This is the slow part by hand, because a single sentence can hold several entities and several relationships at once. Deduplicate as you go: the same character often appears under different names.
Step 4: Map the relationships
Entities are not much use in isolation. Kaelen hates someone, controls a faction, used to serve the queen. Capture those links explicitly, because relationships are what make a world feel connected and what your systems will query at runtime, as in "who in this district has a grudge against the player's patron."
Step 5: Structure the narrative
If your game has authored story, lay it out as structure the engine can follow: arcs, beats, and scenes, with the conditions that gate them, such as "this beat unlocks if the player sided with the smugglers." A node-based flow is the usual shape. This is where static lore becomes something playable.
Step 6: Export to your engine
Finally, turn the structured world into the format your pipeline consumes: an Unreal DataTable, a Unity ScriptableObject, JSON for a custom engine. The point of all the structuring is this moment. One export, and your narrative is data your game can run, not a document a programmer has to translate by hand.
Doing it by hand vs with a tool
All six steps are doable manually, and for a small world they are fine to do in a spreadsheet and a flowchart. The cost is time and consistency: extraction is tedious, relationships are easy to miss, and every change to the lore means redoing the downstream work.
This is the job LoreWeaver Architect automates. You import your existing documents, and it extracts the entities, suggests a schema, maps the relationships it finds, and lets you design the narrative flow on a canvas, then exports to your engine. You keep full editorial control over everything it proposes, and it is free to use. Whether you do it by hand or with a tool, the destination is the same: lore, turned into data your engine can run.
FAQ
How do I convert a story bible into game data? Inventory the lore, define entity types and fields, extract the entities, map their relationships, structure the narrative, and export to your engine's format. The work is turning prose into typed records with explicit links.
Can AI turn my lore into game-ready data? AI is good at the extraction and structuring steps: reading prose and proposing entities, fields, and relationships. Architect uses it for exactly that, with you reviewing and editing what it suggests. AI accelerates the tedious part; you keep control of the result.
What formats can I export narrative data to? Commonly Unreal DataTable, Unity ScriptableObject, and JSON for custom engines. Architect supports these plus custom templates so the output matches your pipeline.
Do I lose creative control if a tool structures my lore? Not if the tool keeps you in the loop. Good tooling proposes structure and lets you approve, edit, or reject every entity and relationship. The creativity stays yours; only the manual formatting is automated.
How is this different from a worldbuilding wiki? A wiki stores lore for people to read. Game-ready data is lore shaped for an engine to execute. A wiki is the source; the structured export is what your game runs. See tools like World Anvil, with AI for the wiki side.
The short version
The distance between a written world and a playable one is structure, and structure is work you can do by hand or hand to a tool. If you want the fast path, Architect turns your lore into engine-ready data and is free to try. Then Director can run it live.