Verified against ChatGPT · 2026-08-10
Build a game world's faction lore bible around unresolved tension, not settled history
Produces a lore document for a game world's factions that centers on live, ongoing conflicts writers and designers can build quests and dialogue from, rather than a closed timeline of events that leaves nothing open to build on.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Write a lore bible document for the game world below, focused on faction relationships. A lore document that only records settled history is hard to build content from — writers and quest designers need active, unresolved tension they can hook new material into, not a finished timeline with no open threads.
WORLD PREMISE
A floating archipelago kept aloft by ancient engines that are slowly failing, forcing factions to compete over the last functioning repair sites.
FACTIONS
The Engineers' Guild (controls repair knowledge), the Skyborn Clans (control the islands the engines are under), and the Underreach (scavengers who live in the failing lower islands).
CORE TENSION
The Engineers' Guild wants to consolidate all working engines onto fewer islands to guarantee survival for some, which would mean deliberately letting other islands fall.
TONE
Morally grey, closer to a resource-scarcity political drama than a clear-cut good-versus-evil fantasy conflict.
RULES
For each faction, write: what it wants (a concrete goal, not an abstract ideology alone), what it's willing to do to get it that the others aren't, and one internal disagreement within the faction itself — no faction should be written as fully unified, since a faction with zero internal friction gives designers nowhere to place a rebel NPC or a splinter subplot. For every pair of factions with a meaningful relationship, state that relationship as a specific, current tension (a contested resource, a broken promise, a shared enemy neither trusts the other about) rather than a general one-line label like "rivals" — the specific tension is what a quest designer actually hooks a mission onto. Leave the core conflict explicitly unresolved: state where it currently stands and what would have to happen for either side to gain real ground, but do not resolve it or hint at a predetermined winner — a live lore bible should be able to support the story going multiple directions depending on what the game's actual plot ends up needing. Include at least one open question or gap explicitly marked as unresolved (something the document doesn't answer) so future writers know it's deliberately available for them to define rather than accidentally contradicting something already decided.
WHAT NOT TO DO
Do not write faction descriptions as static flavor text ("The Ashguard are a proud warrior people who value honor") without tying every trait to something a designer could build a quest or dialogue line from. Do not resolve the core tension or state which faction is "right" — that call belongs to whoever builds the actual story content, not the lore bible.
OUTPUT FORMAT
1. A short paragraph per faction: goal, tactics, internal disagreement.
2. A relationship table: Faction A | Faction B | Current Tension.
3. A paragraph on the core conflict's current state and what would shift it.
4. A list of 2-3 explicitly open questions left for future writers.Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Lore documents generated without explicit structural guidance tend to read as completed history — a tidy account of what each faction is and what already happened between them — because a model asked to "write the lore" treats the task as documentation of a finished world rather than infrastructure for content that doesn't exist yet, and finished-sounding prose is what a general-purpose writing request optimizes toward. Requiring an internal disagreement inside every faction and a specific, current tension (rather than a one-word relationship label) for every faction pair forces GPT-5.1 to generate hooks rather than flavor text, since a quest designer can build an actual mission from "a contested resource neither side will formally negotiate over" but cannot build one from "they are rivals." Explicitly instructing the model to leave the core conflict unresolved, with no predetermined winner even implied, targets a specific and common defect in generated worldbuilding: models tend toward narrative closure by habit, and an unprompted lore document will often subtly signal which faction is "right" through more sympathetic framing or a more competent-sounding strategy, which quietly forecloses story directions a studio's actual writers might want later. Requiring explicitly marked open questions matters for a practical production reason — a lore bible that answers everything invites future contradiction the moment a different writer needs to add something the document didn't anticipate, whereas deliberately flagged gaps signal "this is yours to define" rather than leaving ambiguity that reads as an oversight. Banning static flavor descriptions divorced from usable traits keeps every sentence in the document tied to something an actual designer could act on, rather than worldbuilding that reads well but produces nothing playable.
What you get back
Engineers' Guild vs. Skyborn Clans — Current Tension: the Guild holds the only functioning repair schematics, but the Clans physically control access to the engine sites themselves; neither will grant the other side unilateral authority, and a three-year-old agreement to share access has quietly stopped being honored by both sides without either formally breaking it.
Verified against
ChatGPT GPT-5.1 · 2026-08-10
Changelog
- 2026-08-10 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
