Explore the studio
From a spark
to a seat at the table.
Shape your world, connect its stories, and prepare the next night of play. The council provides structure and ideas; you decide what belongs in your campaign.
01 / The campaign toolkit
Guided creation
The twenty-three stages
Move through the specialists in dependency order. See what is finished, what is waiting on earlier work, and what each stage will add. Independent stages can run concurrently when the server is configured to support it.
Campaign DNA
Tone, themes, inspirations, play style, length, party composition, rules assumptions, safety boundaries and your own preferences. Every expert reads this before it writes anything.
A workspace per specialist
Each expert asks its own focused questions, names what is missing, offers alternatives and explains the trade-offs — rather than one chat box doing twenty-three different jobs.
Approve, edit, compare, regenerate, lock
Control the campaign at the level of single ideas. Regeneration writes a second version beside the first and shows a field-by-field diff, so you can take the new backstory and keep the old motivation. Lock a fact and later work cannot quietly change it.
Three depths
Quick draft, guided build, deep build. The same specialists, asked for different amounts of material.
02 / The campaign toolkit
Campaign intelligence
The shared bible
Approved facts, terminology, themes, boundaries, rules assumptions, characters, factions, secrets and open questions — the campaign's source of truth, and what each expert is handed before it starts.
The knowledge graph
Characters, places, factions, quests, clues, items and events connected by explicit typed relationships, extracted automatically from what each specialist declares. Not just what exists, but how the pieces bear on one another.
The compiler
Review structural findings such as dangling references, mistyped relationships, encounter arithmetic, branch coverage, duplicate names, misfiled entities, unscheduled content, isolated entities and disconnected geography. Inspect the evidence and decide what needs repair before play.
Dependency-aware revision
Change an antagonist, a faction or a historical event and the graph names what is downstream. Repair the clues, motivations and plans actually affected instead of regenerating a world.
Full-text search
Every word of canon, ranked so that typing a canonical id returns that entity rather than the four sessions that mention it.
Ask your campaign
Ask a question about saved campaign material and inspect the quoted sources behind the answer. Start with approved canon, or explicitly include drafts. Answers remain separate from your campaign; changed canon requires a fresh answer. Retrieval is bounded, and you remain the judge of whether a quotation supports an interpretation.
History and branches
Keep entity revisions and compare what changed. Fork a campaign to try something, and compare the two.
03 / The campaign toolkit
Player-centred playability
Backstory weaving
Character goals, fears, bonds and rivals become recurring connections. Spotlight checks help you notice when one character has fewer campaign hooks than the rest.
Faction and antagonist simulation
Important actors get goals, resources, plans and progress clocks. Advance clocks and apply selected consequences to keep the world responding when the party intervenes, negotiates or walks away.
Branching consequences
Prepare outcomes for success, failure, avoidance, alliance, betrayal and the unexpected — and each can move a clock, so a consequence is state rather than a note about state.
Rules-aware encounters
Use the selected supported rules profile and the encounter's intended party level. Review difficulty findings alongside terrain, objectives and special abilities.
Mixed play
Combat, exploration, social, investigation, puzzles and downtime, with checks that compare the generated mix against your brief.
04 / The campaign toolkit
At the table
Session packets
The run sheet: opening situation, beats, NPC cards, stat blocks, clues, encounters, contingencies and likely stopping points.
Rehearsal
What breaks first. Single points of failure read off the graph — a session whose every clue comes from one NPC is one dead NPC away from a halt.
Live mode
A surface you touch while playing. Mark beats as they land, tick clocks when the world moves, reveal a place to the players on their own screen. What you record is what the aftermath reads, so it does not have to guess. Recover unfinished notes in the same tab, then save them to the campaign. Draft recovery clears on sign-out.
Post-session adaptation
Type what happened. It reads that against the plan, proposes what moved, and changes nothing until you confirm.
The prep planner
One ranked queue instead of four lists: what is worth the next hour, weighted by severity, whether it bites at the table, and how soon the session that needs it arrives. Every item shows its arithmetic.
05 / The campaign toolkit
Seeing it
The storyboard
Arrange people, places and events from your campaign alongside notes and drawings. Canon-linked nodes follow the saved entry; your own drawn connections remain separate from canon.
Four layouts
A web ranked by connectedness, a tree of descent and allegiance, an arc along the sessions, or an approximate map informed by travel times between regions.
Findings on the canvas
Continuity problems appear as badges on the nodes they concern, so a hole in the plot shows up as a gap in the picture.
Artwork
Portraits, maps, scenes and banners drawn from what canon says about each thing, through a configured image service. Review the result for visual continuity before keeping it.
06 / The campaign toolkit
Ownership
Player links
A read-only view of exactly what you have revealed, with no account needed. Players can reload the view to see newly revealed material.
Exports
Markdown, a GM bible, player handouts, a portable re-importable bundle, and a Foundry VTT module — with a separate players' copy containing only revealed material.
A portable setup
The application supports local operation with configured model services. A future hosted offering is a deployment plan, not a subscription service promised on this page.
An illustrative table, not a generated campaign
When the party misses the witness.
The miller has disappeared. Your players still have the flour-dusted cloak and a delivery ledger to follow. Trace those leads in the mystery designer, test what happens without the witness, and bring the surviving routes into your session notes.
The tools help you prepare alternatives. Your judgment makes them fit the people around your table.
What it does not do
Quest Council does not answer a player mid-scene or make the Game Master’s decisions. Specialists generate material that you can review, revise and regenerate. Check its rules, pacing and suitability before play.
Artwork depends on a configured image service. Collaboration and invitation management remain deferred; read-only player links are a separate capability. The current specialists use distinct prompts, context and output contracts; separate trained models are not claimed.
Automated checks cover seven structural areas of the evaluation scorecard. They are evidence to inspect, not a verdict on story quality. Blinded human results are not yet published. Read the evaluation method →