Quest Prep
Quest prep is where responsive generation becomes practical. A player posts intent; the GM generates what’s needed; the results become canon; the session runs.
The workflow is always the same: player declares intent, GM checks existing canon, identifies gaps, selects the appropriate system, generates content, updates the wiki.
Step 1: Read the Intent
A player posts something like: “I want to find the assassin who killed my mentor. Looking for anyone in the Thieves Guild who might know. Who’s in? Thinking Thursday.”
Parse this into:
- The goal — find the assassin
- The approach — social investigation through the Thieves Guild
- The implied location — wherever the Thieves Guild operates
- The gap — does the Thieves Guild have an established presence? Do we know anything about this assassin?
Vague intent (“I want to explore”) needs a follow-up question: “What are you hoping to find?” Goal-driven play requires a goal.
Step 2: Check Existing Canon
Before generating anything, review what already exists:
- What’s been established about the relevant factions in the wiki?
- Have other parties visited the relevant location?
- What lore nodes have already been pinned that connect to this goal?
- What faction clocks are running that might intersect?
This step prevents contradictions and often surfaces content that’s already half-generated. The world has been running between sessions. Check what’s there before adding more.
Step 3: Identify the Gaps
What needs to exist for this session to work that doesn’t already exist?
- A location the party hasn’t visited needs generating
- An NPC the player needs to find doesn’t have an established presence
- A dungeon site has been rumored but never realized
Each gap maps to a generation system. A new settlement location needs Ex Novo. A key NPC needs Street Magic. A dungeon needs Ex Umbra. A wilderness route needs Hexmancer.
Step 4: Generate
Run the appropriate system for each gap. The Pandemonium Engine handles the rest — everything you don’t explicitly generate is provisionally real in the Shadow Horizon, waiting to be observed.
For social investigations: generate the key NPC, their faction posts, one lore node they’ll share under the right circumstances, and one they’ll protect at all costs.
For dungeon delves: start with the rumors (what do players think they know), generate the site from those rumors, then seed the site with lore that confirms some rumors and contradicts others.
For wilderness expeditions: seed the discovery pool with content tied to the focal player’s goal, generate the destination site, let travel deliver the seeds.
Step 5: Prep Time by Complexity
| Expedition Type | Prep Time | What You’re Generating |
|---|---|---|
| Revisiting known location | 30 min | NPC reactions, faction clock consequences, new lore |
| New settlement, known region | 1-2 days | Medium generation + key NPCs + hooks |
| New dungeon site | 2-3 days | Rumors → site → ecology → treasures |
| Major new settlement | 3-5 days | Full Ex Novo generation, possibly a building session |
The GM preps after knowing what the players want, not before hoping they’ll bite on a hook. This is why the timelines work — you’re not maintaining a whole world, just generating the specific thing they’re going to.