Worldbuilding Layers and Workflow
Summary
A conceptual model for how the repository is structured, how worldbuilding discovery flows, and when content should move between layers. This document explains the methodology. Read it alongside Agent Instructions and Guiding Principles.
The Three Layers
The repository operates with three distinct content layers. They are not just folder names — they represent different relationships to the world's objective reality.
1. Lore Layer
What it is: Reusable world truths. Information that is stable, objective, and useful across multiple contexts — stories, generated answers, future game systems, further worldbuilding.
What belongs here:
- How WTA arenas work
- How water systems and hydraulic infrastructure operate
- Companion cockatoo social roles and cultural patterns
- Gambling system mechanics
- Animal handling traditions
- Extinction patterns and timelines
- Preserve kitchen workflows and food storage systems
- Any system, institution, or cultural practice that could appear in multiple stories
How lore should be written: Declarative and systemic. Plain, precise language. Focused on mechanisms, behaviours, workflows, constraints, and historical evolution. Not narratively framed. Not emotionally driven. See the prose style guidance below.
What lore is not: Scenes, arguments, emotional experiences, or one character's specific encounter with a system.
Lore is: the repository's canonical model of historical reality, described so it can be retrieved and reused.
2. Story Layer
What it is: Specific traversals through the world. Scenes, interactions, and experiences that emotionalise, contextualise, and compress lore into encounter.
What belongs here:
- The Tinamou yelling at a bookmaker in the arena district
- A tourist discovering an impossible bird call near the reservoir
- A specific argument at the water office
- One worker's bad morning in the preserve kitchen
- Any moment that is about character and pacing rather than world description
How story should be written: Grounded, sensory, emotionally textured, funny where the world is funny. Focused on ordinary people in ordinary situations under recognisable pressure. The correct writing register for the Tinamou story is described in detail in Tinamou Story — Tone and Writing Style Guide.
What story is not: Reusable world truth by default. Lore documentation dressed as drama. Abstract exposition delivered in narrative form. A lower-quality form of lore.
Story is: an experience inside the world — not a description of the world. Stories are traversals through lore substrate. The lore defines what exists; the story experiences it.
3. Interpretation Layer
What it is: Subjective, contextual, and potentially unreliable accounts of how people within the world understand — or misunderstand — objective lore.
What belongs here:
- Museum plaques that are partially or specifically wrong
- Tourist guide scripts (accurate on major sites, wrong on specific objects)
- In-world academic summaries and phone AI descriptions
- Academic papers with contested interpretations (the disagreement itself is canonical)
- Tinamou commentary (emotionally accurate, not always factually reliable)
- Imperial propaganda (self-serving characterisation of political history)
- Popular mythology about lost eras (First Basin as golden age; WTA as civilised order)
- One era's incomplete account of an earlier era
- Reconstructed histories working from fragmentary evidence
Why this layer exists:
The interpretation layer is not a failure — it is the theme. The gap between what existed and how later societies understand it is one of the core tensions of Cassowary World. Ordinary First Basin infrastructure becomes mythologised. Archive halls become temples. Drainage chambers become ritual pools. Maintenance symbols become holy glyphs.
Interpretation-layer material may itself become historically important regardless of factual accuracy. A myth that is wrong about the First Basin but believed for three hundred years is a real historical fact about those three hundred years. Propaganda that misrepresents WTA incorporation is still a document of how that power described itself. Wrong museum labels are evidence of how a later institution read an earlier one.
Concrete examples of interpretation-layer content:
Functional misidentification: A preserve storage facility near the Kati Thanda basin is labelled in museum records as a "Civic Memory Archive." Later archaeology recovers paste-cake residue in the sealed vessels. The misidentification is not false canon — it is documented evidence of how later scholarship misread the site.
Romantic inflation: Post-collapse oral traditions describe the First Basin as an era of extraordinary abundance where the floods always came on time and no one went hungry. The administrative records (where recoverable) show a hydraulic bureaucracy under increasing resource pressure, with intensification, rationing, and coercive labour in its final centuries. Both accounts are canonical. The mythologised version is what people believed; the administrative record is what was happening.
Propaganda: WTA accounts of frontier integration describe the trade guarantee as "bringing stability to ungoverned regions." Regional accounts describe the same process as coercive incorporation into an extractive network. Both framing choices are canonical documents of their position.
Academic disagreement as canon: "Whether Basin-period flood markers functioned as practical water-level calendars or as religious objects remains contested in contemporary scholarship." The markers are canonical objects. The scholarly disagreement is canonical uncertainty. Neither requires resolution to be documented.
The Tinamou's authority limit: The Tinamou was there and remembers. That is a different kind of authority than institutional scholarship — not necessarily superior. The Tinamou may romanticise specific individuals while being accurate about systems. It may remember one incident clearly and generalise from it incorrectly. Its emotional proximity makes it reliable in some ways and unreliable in others.
The Tinamou's emotional attachment to dead workers is more accurate than the museum plaques. The museum plaques are more institutionally authoritative than the Tinamou. Neither is fully correct.
Interpretation is: how people understand the world, not necessarily how the world objectively works. Do not confuse interpretation content with lore content. A museum plaque in a story is not a source of objective world facts. It is a source of information about how that institution, at that moment, understood that object.
The Worldbuilding Workflow
Discovery in this repository follows a four-stage direction. Not a rigid procedure — a natural flow.
Stage 1: Exploration
Generate scenes, ideas, and speculative systems freely. Exploratory content is where the world becomes alive. Emotional resonance is a useful signal — a scene that feels true is worth investigating even if the underlying system is not yet formalised.
Exploratory content in conversations and branches:
- Does not automatically become canon
- May contain exaggeration, emotional framing, or character-specific misunderstanding
- Becomes canon only when extracted into a committed lore document and merged to main
Stage 2: Emotional Filtering
Keep what feels alive, grounded, revealing, or useful. Discard what feels like obligation — scenes that exist to transmit information rather than experience it.
The test: does this moment make the world feel real? Does this detail produce recognition — "yes, that is how it would actually work"?
If yes, it is worth extracting.
Stage 3: Extraction
Promote reusable truths into lore docs. A world system discovered through a scene should live in a lore document where it can be retrieved and recombined with other content.
Promote to lore when:
- The information is reusable across multiple stories or contexts
- It describes a world system, institution, ecology, or cultural pattern
- It explains how things generally work rather than how one moment went
- Multiple future scenes might depend on it
Keep in story when:
- It is scene-specific and depends on character or pacing
- It is one traversal through a system, not a description of the system
- Emotional framing would be misleading if presented as objective lore
- It would only make sense in this particular story moment
Stage 4: Story Construction
Use lore as the substrate for authored experience. The lore layer supports the story layer — the story layer does not need to carry the weight of world exposition.
A scene set in the arena district does not need to explain how arenas work. The lore doc does that. The scene just needs to show one moment inside a working arena, with the lore as invisible support.
Conversations and Branches as the Experimentation Layer
The repository is not the exploration space — it is the canon space.
Conversations are the experimentation layer. Scenes written in chat, speculative systems explored through dialogue, contradictory ideas tested — none of this is canon automatically. It can become canon when extracted into a committed lore document and merged to main.
Branches and pull requests are the iteration layer. Work in progress, proposed additions, contested ideas live in branches until they are reviewed and merged.
Main branch is canon. What is committed to main is authoritative, reusable, and the baseline for all future work. If canon changes, it is updated directly. There is no formal intermediate state between "conversation" and "canon."
The implication: no special metadata flags or stability tiers are needed in the repository itself. Git handles the experimentation system. Documents in the repository are canonical by being there.
In-world uncertainty is still valid canon
The world itself can contain genuine uncertainty — contested archaeology, competing scholarly interpretations, mythologised accounts of lost eras, incomplete records. This is not authorial uncertainty about what the world is. It is a real feature of the world, documented as canon.
Example: a lore document can state that "modern scholars disagree about whether Basin flood shrines originated as maintenance offices." That disagreement is itself canonical. The uncertainty lives inside the world, not inside the authoring process.
Prose Style Guidance
Lore documents and story documents have different goals and different writing modes. The guidance below applies specifically to each type. It is not a single unified writing style applied uniformly across the repository.
The short version: lore should read like an encyclopedia entry about a real historical institution. Story should read like an account of being inside that institution on a specific day.
Lore documents: declarative and systemic
Lore documents should prioritise systems, behaviours, relationships, workflows, ecology, and constraints. They should not read like scenes, cinematic descriptions, or emotional accounts.
Bad — cinematic prose in a lore doc:
The crowd roared as the beast emerged from shadow, its eyes catching the torchlight as the arena fell into terrified silence.
Good — declarative system description:
WTA arenas commonly used concealed-release predator gates in waterhole layouts to increase crowd tension and betting volatility. The delay between gate opening and predator emergence was a managed part of the event structure.
Bad — emotionally narrated lore:
Workers trudged wearily through the drying yards, their backs bent under the weight of seasons of labour, the smell of sweet fruit hanging heavy in the air like a memory of better days.
Good — workflow and mechanism:
Drying yards operated on rotation schedules keyed to ambient temperature and humidity. Output in peak season was monitored by a dedicated record-keeper responsible for identifying spoilage before it spread through the batch.
Bad — abstract fantasy narration:
The hydraulic administrator reviewed the sacred obligations of water stewardship, connecting the people to the life-giving channels of their ancestors.
Good — practical institutional description:
Water allocation in the First Basin period was administered through a desk-level record system using clay tags. Workers reported leaks and submitted repair claims through the same office that managed water turn schedules.
The test: could this sentence appear in an encyclopedia entry about a real historical institution? If yes, it is probably correct lore style.
Story documents: grounded and experiential
Story documents are not lore documents with atmosphere added. They are a different kind of writing serving a different purpose. They should be sensory, grounded, funny where the world is funny, and emotionally textured.
Story writing guidance:
- Do describe specific objects, jobs, smells, sounds, queues, arguments, and ordinary working conditions
- Do allow humour, confusion, emotional framing, and sensory detail
- Do not write dry encyclopedia entries — that is not what story documents are for
- Do not write cinematic fantasy narration, mythic prose, or theatrical exposition
- Do not bury world systems in exposition when a scene can show a system under pressure instead
The intended register is closer to Pompeii archaeology tourism than to Tolkien narration: specific, lived-in, practical, slightly chaotic, funny in the way that ordinary life under institutional pressure is funny.
The full register guidance for the Tinamou story is in Tinamou Story — Tone and Writing Style Guide.
Retrieval-Oriented Structure
The repository is used as a knowledge base as well as a creative document. Future tooling will retrieve, cross-reference, and recombine content. Structure should support this.
Prefer smaller, focused lore documents over large mixed-topic files.
A document focused on WTA arena layouts retrieves better than a general history of WTA entertainment. The focused document can be used in responses about specific arena configurations without pulling in unrelated content.
Prefer:
wta-arena-culture-and-animal-spectacle.md— arena-specificcompanion-cockatoos.md— cockatoo-specificfood-preservation-and-storage-systems.md— preservation-specific
Over:
wta-era-culture.md— everything WTA in one fileanimal-relationships.md— all animals, all eras, indiscriminate
Atomic Notes are the most retrievable unit. Each note should state one fact, one relationship, or one constraint. A reader — human or automated — should be able to understand the note without reading the whole document.
Discourage:
- Reusable truths buried inside narrative scenes
- Multi-topic prose files mixing unrelated systems
- Emotionally written lore entries that require context to interpret
Encourage:
- Strong metadata with accurate topics, time periods, and regions
- Cross-links to related documents
- Open Questions sections for genuine uncertainty
Regional Variation and Historical Messiness
Cassowary civilisation spans a large and ecologically diverse geography across Sahul. Regional variation is a feature of realistic historical modelling, not a problem to be smoothed over. Lore documents should reflect this.
Sources of regional variation
Ecological diversity: Coastal, riverine, inland, arid, upland, and rainforest regions have different food systems, different available fauna, different water availability, different climate pressures, and different relationships to the preserve circuit. A coastal city near riverine trade routes has a different daily life than an inland arid-margin city that is marginal to the WTA network.
Historical exposure: Different regions experienced the Collapse Era differently. Some were deeply embedded in haulage-dependent infrastructure and collapsed hard. Others were peripheral to that infrastructure and continued at lower institutional thresholds. Recovery paths differ. What a city's ruins look like, what its population remembers, and what they mythologise depends on which collapse experience they had.
Trade integration: Core WTA nodes, peripheral participants, and non-integrated regions operate under different economic conditions. A city at a major exchange node has different social structures from a frontier outpost that receives WTA guarantees but sits at the edge of the network.
Religious and cultural drift: Regional flood deities, local preservation-associated cults, trade-route protective entities, and ancestor practices vary by ecology and history. A coastal city's religion relates to river floods and sea access differently than an inland arid-margin city's religion relates to desert-edge water sources.
Arena culture variation: Major urban arenas with permanent facilities, professional handlers, and formal gambling infrastructure differ substantially from frontier outpost versions with less infrastructure, fewer animals, and improvised arrangements.
What coherence means
The WTA framework creates institutional coherence, not uniformity. Coherence comes from: shared WTA route guarantees, compatible administrative language, common seal-verification conventions, recognisable pottery standards. Diversity comes from what those shared institutions are managing in each specific ecology, and from the different histories each region brings to WTA participation.
A pottery guild mark has the same meaning in different cities because the institutional convention is shared. What that pottery guild makes, who its members are, what their lives look like, and what they eat for lunch will differ significantly between a coastal ceramics district and an inland arid-region workshop.
Guidance for lore writing
- Examples in lore documents should be illustrative, not universally prescriptive. "A typical WTA-era city" is not the same as "all WTA-era cities."
- When specific examples appear, note where they apply (Calisander, inland route nodes, coastal districts) rather than implying universal application.
- Open questions about regional variation are valuable to record. "How does this system differ in frontier outpost contexts?" is a legitimate open question.
- Do not accidentally assert that any described institution, custom, or social arrangement is universal across the WTA network.
Common Failure Modes
Encyclopaedia sludge: Lore docs that accumulate too much information per topic, becoming slow to read and difficult to retrieve from. Fix: split into focused documents.
Reusable truths trapped in scenes: A scene reveals something important about how the world works, but the fact never makes it into a lore doc. Fix: extract after the scene is written.
Lore docs becoming stories: A lore doc starts narrating events cinematically. Fix: convert back to declarative structure; move emotional material to story docs.
Stories becoming dry wiki text: A scene becomes a list of facts about the world. Fix: return to character, moment, and specific ordinary pressure.
Interpretation mistaken for lore: A museum plaque or a character's account of history is treated as objective world fact. Fix: check who is speaking and what their epistemic position is.
Over-mythologising: A practical system — a jar store, a water office, a drying yard — gets described in cosmic or ceremonial language because it is ancient. Fix: describe it as it was to the people who used it daily, not as it appears to later romanticisers.
Assuming homogeneous empire: A single example of WTA-era practice is treated as universal. Fix: scope the example to where it applies; note regional variation where it is likely; record open questions about how systems differ in different ecologies and integration levels.