Long-term content strategy workshop: 300-world generator model, cultural ingredients menu, three-system NPC architecture (9 patterns x 6 motivations), Sacred/Profane/Middle Kingdom framework. 9 agents across 4 rounds plus lead interview establishing the production path from hand-authored Sova to generated 300 worlds. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
44 KiB
Workshop Round 4 — Miri (Worldbuilder / Setting Designer)
Responding to the lead's decisions. Reframing for 300-world generator-input scale.
What Changes at 300 Worlds
Everything I proposed in Round 3 was designed for 8-10 hand-authored regions. At 300 worlds, the paradigm inverts:
| Round 3 assumption | Round 4 reality |
|---|---|
| Regional brief is documentation a human reads | Regional brief is a parameter set a generator consumes |
| Cultural groups are a fixed taxonomy | Cultural groups are assembled from an ingredients menu |
| Naming pools are hand-curated lists | Naming is algorithmic — phonetic rules, not word lists |
| Social norms are prose descriptions | Social norms are numeric parameters the dialogue system reads |
| 8-10 briefs, each hand-authored | 8-12 ingredient definitions that combine into thousands of unique societies |
| Quality from expert review | Quality from constraint specification — the generator can't produce garbage |
| Each region needs a worldbuilder | The worldbuilder designs the menu; generators cook the meals |
The Krenn System brief from Rounds 1-2 is now a proof-of-concept output — one meal from the kitchen. The real product is the recipe book + the ingredient pantry + the cooking rules.
1. The Ingredients Menu
The lead is right: pick-and-combine is better than fixed taxonomy. Here's the menu, designed as generator input.
Ingredient Category A: Heritage Roots
Heritage roots are phonetic and social flavor profiles. Each root is a parameter bundle the generator uses to produce naming, social dynamics, and cultural texture. The player recognizes the FEEL without being able to say "that's Finnish."
The roots are named for their social signature, not their Earth origin. A developer might think "Frost is Nordic-inspired" but the generator just knows "compact phonetics, social reserve, patience-based trust."
I'm proposing 10 roots. A society picks 1-3 and blends them with drift weighting.
| Root | Phonetic Signature | Social Signature | Trust Model | Authority Attitude | Player Feels |
|---|---|---|---|---|---|
| Frost | Compact, consonant-clusters, 1-2 syllables, hard stops (k,t,v,s) | Reserved, privacy-first, stoic | Patience — time + shared labor | Formal distance — cooperate without volunteering | "Northern, quiet, self-reliant" |
| Tide | Flowing, vowel-rich, 2-3 syllables, soft consonants (l,n,m,r) | Social, group-oriented, expressive | Hospitality — sharing food, sharing space | Personal — authority must earn affection | "Coastal, warm, community-spirited" |
| Iron | Heavy consonants, rhythmic, 1-2 syllables, gutturals (g,kh,r) | Communal solidarity, endurance, labor-proud | Collective — trust the group, not individuals | Suspicious — institutions are outsiders | "Industrial, tough, workers' culture" |
| Spice | Layered, multi-syllable, precise stress patterns, sibilants | Extended kinship, layered formality, spiritual undertones | Kinship — blood and marriage connections | Respectful-formal — hierarchy acknowledged, not resented | "Elaborate, family-oriented, ritualistic" |
| Jade | Precise, tonal tendency, short but varied, clean consonants | Hierarchical, honor-aware, aesthetic-conscious | Competence — demonstrated skill earns respect | Institutional — authority is legitimate if earned | "Precise, beautiful, structured" |
| Dust | Rhythmic, open vowels, 2-3 syllables, nasals (n,m,ng) | Communal decision-making, oral tradition, trade-savvy | Witness — trust requires public demonstration | Collective — the community outranks the institution | "Rhythmic, communal, oral-tradition-rich" |
| Vine | Warm, 2-3 syllables, soft stops, rolled consonants | Class-conscious, family honor, emotionally warm | Family — trust through recognized kinship or patronage | Stratified — authority respected based on social class | "Warm, class-aware, passionate" |
| Salt | Pragmatic, clipped, 1-2 syllables, dental consonants (d,t,th) | Individualist, contract-based, commercial | Transaction — trust through fair dealing | Pragmatic — authority is a service, evaluated by results | "Merchant, practical, no-nonsense" |
| Stone | Steady, mid-length, balanced consonant/vowel, laterals (l,r) | Enduring, traditional, land-connected, seasonal | Tenure — trust through long presence | Patriarchal/Matriarchal — elders hold authority by custom | "Rooted, agricultural, slow-moving" |
| Arc | Sharp, variable length, fricatives (s,z,f,v), unusual combinations | Intellectual, cosmopolitan, debate-oriented | Argument — trust through demonstrated reasoning | Meritocratic — authority must justify itself intellectually | "Sharp, academic, argumentative" |
How roots blend
A society picks 1-3 roots with blend weights summing to 1.0. The generator blends parameters:
# Example: Krenn System
heritage:
primary: frost # weight: 0.55
secondary: salt # weight: 0.30
tertiary: iron # weight: 0.15
drift_stage: mature # 180 years
The generator produces:
- Naming phonetics: 55% Frost rules (compact, consonant-cluster) + 30% Salt rules (clipped, dental) + 15% Iron (heavy, rhythmic) = short, hard names with practical feel. This IS Krenn naming — we just derived it from parameters instead of hand-authoring it.
- Social dynamics: 55% reserve + 30% pragmatism + 15% solidarity = mind-your-business with transactional undertones and labor solidarity. This IS Krenn culture.
- Trust model: Dominant = patience (Frost) modified by fair-dealing expectation (Salt) = trust earned through time + demonstrated reliability.
Drift adjustment: The drift stage modifies how blended the roots are. At "pioneer" (0-50yr), the primary root dominates strongly (weight 0.8+). At "mature" (150-300yr), the roots are more evenly blended and novel elements emerge. At "ancient" (300+), the heritage is substrate — barely detectable, mostly novel.
drift_modifiers:
pioneer: # primary ×1.4, secondary ×0.4, tertiary ×0.2, novel: 0.0
crystallizing: # primary ×1.1, secondary ×0.7, tertiary ×0.5, novel: 0.1
mature: # primary ×0.9, secondary ×0.8, tertiary ×0.7, novel: 0.3
ancient: # primary ×0.5, secondary ×0.5, tertiary ×0.5, novel: 0.8
The "novel" factor represents cultural elements that have no heritage origin — they emerged from local conditions. Krenn's first-name-primary convention is novel: no founding culture used it, but it emerged from logistics-class social compression. At 300 worlds, the generator needs to introduce novel elements proportional to drift stage.
Ingredient Category B: Settlement Motivation
What brought people here shapes what they value. Each motivation provides defaults for social structure, institutional attitude, and what the community considers "success."
| Motivation | Social Structure Default | Authority Attitude Modifier | What "Success" Means |
|---|---|---|---|
| Economic-extractive | Class hierarchy (owners/workers) | Varies by owner: corporate = compliant, independent = hostile | Profit. Getting out. Sending money home. |
| Economic-trade | Merchant networks, guilds, competitive | Pragmatic — regulation is cost of business | Volume. Reputation. Connections. |
| Economic-agricultural | Land-based hierarchy, generational | Patriarchal/matriarchal elders | Land. Harvest. Family continuity. |
| Ideological-political | Factional, democratic or authoritarian | Depends on ideology — could be deeply institutional or anti-institutional | Advancing the cause. Purity of vision. |
| Ideological-academic | Competence hierarchy, flat social | Meritocratic — authority must be earned through expertise | Knowledge. Discovery. Recognition. |
| Institutional-military | Chain of command, strict hierarchy | Total — institution IS authority | Order. Duty. Security. |
| Institutional-administrative | Bureaucratic layers, credentialed | Procedural — authority through process, not person | Position. Influence. Legacy. |
| Refugee | Trauma-bonded, informal, fluid | Distrustful — institutions failed them before | Safety. Stability. Being left alone. |
| Frontier-adventurist | Chaotic, reputation-based | Absent or hostile — came here to escape rules | Freedom. Self-determination. The next thing. |
Note on the lead's point: "What a society DOESN'T care about is as defining as what it does." A frontier settlement with no ideological motivation has no philosophical framework for its grey economy — people smuggle because they need to, not because they believe in anything. That absence IS the character. The generator should be able to leave motivation categories EMPTY, and the absence produces a society without that dimension.
Ingredient Category C: Economic Function
What the settlement DOES now. May differ from original motivation.
| Function | Daily Rhythm | Gathering Pattern | Investigation Texture |
|---|---|---|---|
| Extraction (mining, resource) | Shift-based, dangerous | Canteen, shift change, safety briefings | Physical evidence, equipment access, safety records |
| Logistics (trade hub, freight) | Shift-based, routine | Bar, dock transitions, manifest processing | Manifest discrepancies, schedule anomalies, cargo tracking |
| Manufacturing | Shift-based, specialized | Guild hall, workshop, union meetings | Process knowledge, supply chain, quality records |
| Agriculture | Seasonal, weather-dependent | Market days, harvest festivals, family meals | Land records, water rights, seed stocks, generational grudges |
| Services (hospitality, maintenance) | Client-based, irregular | Public spaces, client premises | Client records, service schedules, reputation networks |
| Research | Project-based, irregular | Lab, seminar, common room | Research logs, peer review, grant records, experiment data |
| Military/Security | Watch-based, alert-cycling | Barracks, briefing room, mess hall | Chain of command, incident reports, patrol logs |
| Administrative | Office-hours, meeting-driven | Office, corridor politics, formal dinners | Paper trails, meeting minutes, budget records |
| Transit (gateway, waystation) | Continuous, transient | Arrival/departure lounges, temporary bars | Passenger manifests, transit records, transient contacts |
| Mixed/Diversified | Varies by sector | Multiple gathering types | Multiple investigation vectors |
Ingredient Category D: Economic Pressure
The lead's addition: what makes extra income tempting. This determines the MORAL TEXTURE of the grey economy — not just what's smuggled, but WHY people participate.
| Pressure Type | Why People Smuggle | Moral Texture | Investigation Emotional Weight |
|---|---|---|---|
| Tight margin | Enough to survive, not to thrive. Extra income = breathing room. | Sympathetic — "we're not greedy, we're trapped" | Busting this feels like punishing poverty |
| Debt trap | Owe more than you earn. Extra income = freedom from debt. | Desperate — "I have no choice" | Busting this feels cruel unless you address the debt |
| Status competition | Enough to survive, not enough to compete socially. Extra income = social mobility. | Ambivalent — "I want what they have" | Busting this feels like enforcing class boundaries |
| Survival gap | Basic needs unmet. Extra income = food, medicine, heat. | Urgent — "people will die without this" | Busting this is nearly unconscionable |
| Opportunity disparity | Legal work pays; illegal pays 10x. Extra income = rational choice. | Calculating — "it's just business" | Busting this feels like the system working correctly... uncomfortably |
| Prohibition economy | The desired thing is banned. Extra income = access to the forbidden. | Ideological — "the ban is wrong" | Busting this forces the player to evaluate the ban |
| Generational extraction | Wealth flows out of the community to distant owners. Extra income = keeping what's yours. | Righteous — "they're stealing from us first" | Busting this feels like being the owner's enforcer |
Krenn is "tight margin" + "prohibition economy" — workers need extra income AND lattice components are banned. The double pressure creates the specific moral texture where the ring is both economically rational and ideologically defensible.
At 300 worlds, the generator picks 1-2 pressure types per society. The combination determines the grey economy flavor, which determines the contraband type, which determines the investigation's moral weight.
Ingredient Category E: Drift Stage
Already defined above. Four stages: pioneer (0-50yr), crystallizing (50-150yr), mature (150-300yr), ancient (300+). Affects heritage blending, novel element emergence, and institutional calcification.
Ingredient Category F: What's Missing (The Absence Ingredients)
The lead said "What a society DOESN'T care about is as defining as what it does." The generator should track what's ABSENT:
| Absence | What It Produces | Example |
|---|---|---|
| No heritage consciousness | People don't think about where they came from. Names are functional, not cultural. No food traditions, no festivals, no substrate words. | Transit hub populated by transients — nobody is "from here" |
| No economic pressure | Nobody needs extra income. The grey economy is about power or ideology, not money. | Wealthy core world where smuggling is a game, not a necessity |
| No institutional authority | No Commission presence, no Assembly representation. Self-governing by default or by chaos. | Frontier beyond the gate network's administrative reach |
| No community bonds | People don't know each other. No social fabric to protect or tear apart. | Transient gateway station with high population turnover |
| No ideological framework | People don't have a philosophy about their situation. They just survive. | Resource extraction colony where workers are too exhausted to think |
The generator should allow NULL values for any ingredient category. The NULL is itself a parameter that shapes the society.
2. Mechanical Parameter Matrix
The ingredients menu is the human-facing design tool. Underneath, the generator needs a structured parameter matrix that every game system reads.
The Society Profile (Generator Output)
When the generator assembles a society from ingredients, it produces a profile. This profile is what I called the "regional brief" in Rounds 1-3, but now it's machine-generated, not hand-authored.
# Generated society profile
society_id: krenn-sova-transit
world_id: krenn-system
# Ingredients (input)
heritage:
roots: [{id: frost, weight: 0.55}, {id: salt, weight: 0.30}, {id: iron, weight: 0.15}]
drift_stage: mature
drift_years: 180
settlement_motivation: economic-trade
economic_function: logistics
economic_pressure: [tight-margin, prohibition-economy]
corporate_presence: significant # Sova Logistics Consortium
philosophical_alignment: null # absent — no ideology, pure pragmatism
# Naming parameters (derived from heritage roots + drift)
naming:
phonetics:
syllable_range: [1, 2]
consonant_preference: [k, t, v, s, r, d, n, l, p]
vowel_preference: [a, e, o]
cluster_frequency: 0.3 # how often consonant clusters appear
suffix_pool: [-an, -en, -ik, -ar, -ev, -mm, -sk, -ur, -ak]
compound_rate: 0.1 # how often compound names appear
social_convention: first-name-primary
authority_naming: title-plus-surname
formality_threshold: 0.7 # how familiar before first-name is used (0=immediate, 1=never)
# Social parameters (derived from heritage + motivation + function)
social:
privacy_level: 0.8 # 0=everything public, 1=everything private
solidarity_type: labor-informal # how people bond
expression_style: contained # emotional display
gathering_primary: bar # primary social site type
gathering_trigger: shift-end # what triggers gathering
daily_rhythm: shift-based
# Access tier parameters (derived from heritage + motivation)
access_tiers:
public:
topics: [environment, work-complaints, sport, news]
entry_cost: 0 # automatic
peer:
topics: [workplace-gossip, mild-opinions, npc-observations]
entry_cost: 30 # days of consistent non-threatening presence
mechanism: shared-time
insider:
topics: [income-sources, grey-economy, personal-troubles, secrets]
entry_cost: 90 # requires vouching + shared risk
mechanism: demonstrated-risk-sharing
authority:
topics: [formal-answers, minimal-disclosure]
entry_cost: 0 # automatic for institutional NPCs
demeanor: cooperative-unhelpful
hostile:
trigger: [trust-violation, cultural-norm-breach, aggressive-questioning]
expression: withdrawal # vs. confrontation, silence, sabotage
# Trust parameters
trust:
building_mechanism: patience # what earns trust
building_rate: 0.015 # trust points per shared interaction
decay_rate: 0.005 # trust decay per day of absence
betrayal_cost: -0.8 # trust lost on betrayal (0 to -1)
recovery_possible: true
recovery_rate: 0.005 # slower than building
# Authority parameters
authority:
attitude: formal-unhelpful
institutional_presence: commission # which institution is local
compliance_level: 0.6 # how much the community actually complies
reporting_norm: never # cultural norm around reporting to authority
# Investigation texture
investigation:
primary_style: patience-observation # how the detective solves things here
secondary_style: manifest-analysis
noise_floor: cultural-silence # what creates false negatives
signal_type: routine-deviation # what reveals truth
# Grey economy
grey_economy:
structure: cooperative-ring
contraband_primary: lattice-components
contraband_secondary: [medical-grade, severance-tech]
moral_frame: access-denied-by-regulation
participant_motivation: economic-need-plus-ideological-justification
# Environmental texture
environment:
substrate_word_density: 0.12 # fraction of speech with heritage words
sign_register: terse-functional
food_culture_strength: 0.6 # how much heritage food survives
graffiti_style: sparse-operational
music_heritage: 0.1 # almost none
color_palette: [industrial-white, warm-amber, emergency-red, cool-gray]
What each game system reads
| System | Parameters Read | Purpose |
|---|---|---|
| NPC Name Generator | naming.* |
Produces names that sound right for this society |
| Dialogue Selection | access_tiers.*, trust.*, social.privacy_level |
Filters which lines an NPC will say to the PC |
| Monologue Generation | investigation.*, environment.substrate_word_density |
Shapes how the PC's internal voice reacts |
| Routine Generator | social.daily_rhythm, social.gathering_trigger, economic_function |
Creates NPC schedules that fit the culture |
| Storyteller | trust.building_rate, access_tiers.peer.entry_cost, investigation.primary_style |
Paces revelations based on cultural trust dynamics |
| Environmental Text Gen | environment.*, naming.sign_register |
Produces signs, menus, graffiti, tickers that feel right |
| Contraband System | grey_economy.*, economic_pressure |
Determines what's being smuggled and why |
| Triangle Generator | social.solidarity_type, grey_economy.structure |
Creates culturally appropriate conflict patterns |
Validation: Can this reproduce Krenn?
If I feed the Krenn ingredients into this system, does the output match what I hand-authored in Rounds 1-2?
| Hand-authored (Round 1-2) | Generator would produce | Match? |
|---|---|---|
| Names: Kael, Voss, Drin, Renn, Tav | Short, consonant-heavy, 1-2 syllable names | Yes — same phonetic space |
| "Mind your business" ethos | privacy_level: 0.8, authority.attitude: formal-unhelpful |
Yes — produces same social dynamics |
| Trust through shared labor over months | trust.building_mechanism: patience, access_tiers.peer.entry_cost: 30 |
Yes — same timeline |
| Heritage food: leib, supp, kalaa | food_culture_strength: 0.6 + Frost substrate vocabulary |
Approximately — generator would produce different specific words but same density |
| Graffiti: sparse, operational | graffiti_style: sparse-operational |
Yes |
| Investigation: patience + manifest analysis | investigation.primary_style: patience-observation |
Yes |
The generator can't reproduce the EXACT words (leib, supp, kylm) because those are authored instances, not generated outputs. But it produces the same density, the same register, the same feel. At 300 worlds, that's the right fidelity level. The specific words don't need to be hand-picked for every world — a phonetic generator that knows "Frost substrate + mature drift" can produce plausible heritage words.
Exception: Tier 1 content. The hand-authored v0.1 district (Sova) keeps its specific words, specific names, specific environmental text. These are the exemplar — the proof that the generator's output space contains quality content. For all other worlds, the generator operates within the parameter space defined by the ingredients.
3. NPC Pattern + Motivation Composition
Pattern definitions (worldbuilding perspective)
Each thematic pattern has a WORLDBUILDING function — what role does this person play in the community's story?
| Pattern | Worldbuilding Function | Content Tier Requirement | Per-District Count |
|---|---|---|---|
| FRIEND | The person the conspiracy hurts most through the player's relationship to them. Emotional core. | Tier 1 (hand-authored) | 1 per archetype (minimum 2) |
| MIRROR | Shows the player an alternate version of themselves. Same position, different choice. | Tier 1-2 | 0-1 |
| ANCHOR | Makes the place feel like home. Loss = displacement. | Tier 2+ | 1-2 |
| GHOST | Absent but referenced. Shapes events without being present. | Tier 2-3 | 0-2 |
| CATALYST | Their arrival or action triggers change. The disruption. | Tier 2+ | 1-2 |
| THRESHOLD | Controls access to a space, a group, or information. | Tier 2-3 | 1-3 |
| REMNANT | Carries memory of how things were before. Living history. | Tier 2-3 | 0-2 |
| SYSTEM | Embodies the institution. Their behavior IS policy. | Tier 2-3 | 1-3 |
| NOBODY | Background. Becomes foreground through player attention. | Tier 3 (starting) | 30-50% of population |
Functional motivation definitions
| Motivation | What They're Doing | Knowledge State | Mechanical Function |
|---|---|---|---|
| HANDLER | Managing operations | Knows the structure | Provides operational intelligence if accessed |
| WITNESS | Saw something | Knows a specific event | Provides evidence if trusted/pressured |
| TURNCOAT | Switching sides | Knows both sides | Provides access to the other side's information |
| CIVILIAN | Living their life | Knows nothing conspiratorial | Provides noise floor, social texture, moral weight |
| OPERATOR | Performing operations | Knows their piece of the operation | Observable during operations, can be followed |
| SKEPTIC | Doubting the system they're in | Knows something is wrong, not what | Provides leads through questions they're asking |
Composition rules for generators
Rule 1: FRIEND composition determines the emotional core.
The FRIEND's functional motivation defines what kind of betrayal/contradiction the player encounters:
| FRIEND + | Contradiction Type | Emotional Payload |
|---|---|---|
| HANDLER | Your friend is running the operation you're investigating/working for | Power betrayal — "you were using me" |
| WITNESS | Your friend knows the truth and is hiding it from you | Protective betrayal — "you were shielding me" |
| TURNCOAT | Your friend is switching to the other side | Loyalty betrayal — "you're abandoning us" |
| CIVILIAN | Your friend has nothing to do with the conspiracy but is being destroyed by it | Collateral damage — "you're an innocent caught in this" |
| OPERATOR | Your friend is actively performing the operations you're discovering | Identity betrayal — "I didn't know who you really were" |
| SKEPTIC | Your friend is questioning the same things you are, from the opposite side | Parallel discovery — "we're both finding out the truth, but it means opposite things to us" |
Kael in Sova is FRIEND + OPERATOR (active ring member the smuggler discovers in B-7). Sera is FRIEND + WITNESS (sitting on evidence to protect Naia). The generator picks from these combinations to produce different emotional cores per playthrough.
Rule 2: NOBODY composition determines what they become.
A NOBODY starts as Tier 3 with a functional motivation. When the player invests attention, the NOBODY promotes:
| NOBODY + starts as | Player attention reveals | Promotes to |
|---|---|---|
| CIVILIAN | A person with mundane problems that resonate | ANCHOR (if the player keeps engaging) |
| WITNESS | Someone who saw something | CATALYST (their knowledge changes the situation) |
| OPERATOR | Someone hiding in plain sight | MIRROR or SYSTEM (depending on what they're operating) |
| SKEPTIC | Someone asking questions the player is also asking | Can become an ally or informant |
| HANDLER | The real power was the background character all along | Major revelation — the generator should flag this as rare (5-10% of NOBODYs) |
| TURNCOAT | Someone quietly changing allegiance | Provides an unexpected ally or enemy |
The NOBODY → something promotion is the "evolving relationships" mechanic. The storyteller monitors which NOBODYs the player interacts with. After N interactions, the storyteller can promote a NOBODY to a higher pattern. This is emergent narrative — the player's attention creates the story.
At 300-world scale: the generator pre-assigns hidden motivations to NOBODYs. Most (70%) are CIVILIAN and stay NOBODY forever. Some (20%) have motivations that activate on attention. A few (10%) have load-bearing motivations that could reshape the investigation if discovered.
Rule 3: SYSTEM composition reveals institutional character.
A SYSTEM NPC's motivation tells you what the INSTITUTION is really doing:
| SYSTEM + | What the institution is doing | Investigation implication |
|---|---|---|
| HANDLER | The institution is managing operations — possibly the grey economy it's supposed to prevent | Institutional corruption |
| WITNESS | The institution knows but isn't acting | Institutional complicity |
| TURNCOAT | The institution is changing policy/allegiance | Institutional instability |
| CIVILIAN | The institution is just... there. Bureaucratic inertia. | Red herring — the institution isn't the story |
| OPERATOR | The institution is actively performing the thing it claims to oppose | Deep corruption |
| SKEPTIC | The institution doubts itself | Reform opportunity |
Rule 4: Distribution constraints for generators.
Per district (30-40 NPCs):
pattern_distribution:
FRIEND: {count: [2, 4], per_archetype: 1, tier: 1}
MIRROR: {count: [0, 1], tier: [1, 2]}
ANCHOR: {count: [1, 3], tier: [2, 3]}
GHOST: {count: [0, 2], tier: [2, 3]}
CATALYST: {count: [1, 2], tier: [2, 3]}
THRESHOLD: {count: [1, 3], tier: [2, 3]}
REMNANT: {count: [0, 2], tier: [2, 3]}
SYSTEM: {count: [1, 3], tier: [2, 3]}
NOBODY: {count: [12, 25], tier: 3} # 30-60% of total
motivation_distribution:
HANDLER: {count: [1, 3]}
WITNESS: {count: [2, 5]}
TURNCOAT: {count: [1, 3]}
CIVILIAN: {count: [10, 20]} # majority
OPERATOR: {count: [2, 5]}
SKEPTIC: {count: [1, 4]}
# Cross-constraints
forbidden:
- pattern: GHOST, motivation: THRESHOLD # can't guard if absent
exception: "unless the ghost's absence IS the barrier"
rare: # generator flags for review
- pattern: NOBODY, motivation: HANDLER # max 1 per district
- pattern: FRIEND, motivation: CIVILIAN # requires strong collateral-damage writing
Cultural modifiers on pattern distribution
The society profile modifies which patterns and motivations are common:
| Society Parameter | Pattern Effect | Motivation Effect |
|---|---|---|
solidarity_type: labor-informal |
More ANCHORs (community bonds) | More CIVILIANs (solidarity protects non-participants) |
solidarity_type: family |
More THRESHOLDs (family gatekeepers) | More WITNESSes (family members see everything) |
authority.attitude: hostile |
More SYSTEMs (institution is visible antagonist) | More SKEPTICs (community doubts authority) |
authority.attitude: absent |
Fewer SYSTEMs | More HANDLERs (someone fills the power vacuum) |
social.privacy_level: high |
More NOBODYs (people fade into background) | Fewer WITNESSes (people don't see because they don't look) |
social.privacy_level: low |
Fewer NOBODYs (everyone is visible) | More WITNESSes (everyone sees everything) |
drift_stage: pioneer |
More REMNANTs (founders still alive) | More OPERATORs (everyone is building/doing) |
drift_stage: ancient |
More GHOSTs (referenced history) | More CIVILIANs (society is calcified, few active operators) |
This means: a Frost/Salt/Iron society (Krenn-like) with high privacy and labor solidarity NATURALLY produces more NOBODYs and CIVILIANs than a Tide/Vine society with low privacy and family solidarity, which NATURALLY produces more WITNESSes and THRESHOLDs. The cultural parameters shape the NPC population without explicit per-society rules.
4. Eight PC Archetypes
The eight, defined as relationships to the world's systems
Each archetype is a POSITION in a community, not a character class. The generator needs to know what each position looks like, what access it provides, and how NPCs react to it.
| # | Archetype | Relationship to Systems | Starting Access | Starting Knowledge | Moral Position |
|---|---|---|---|---|---|
| 1 | Insider | Embedded in community AND grey economy | peer/insider with community, insider with ring | Knows the operation, doesn't know it's being investigated | "I'm part of this. What does that make me?" |
| 2 | Investigator | Institutional authority investigating | authority with institutions, outsider with community | Knows something is wrong, doesn't know what | "I'm here to find the truth. What if the truth is worse?" |
| 3 | Newcomer | Just arrived. No position yet. | public everywhere | Knows almost nothing. Fresh perception. | "I don't understand this place yet. What am I walking into?" |
| 4 | Operator | Runs the grey economy | insider/authority with ring, peer with community | Knows the full operation | "I built this. What happens when it falls apart?" |
| 5 | Authority | Institutional power (governance, enforcement) | authority everywhere, hostile from grey economy | Knows institutional structure, not street-level reality | "I represent the system. Is the system worth representing?" |
| 6 | Drifter | Between communities. Transient. | public everywhere, but quick peer in transient spaces | Knows patterns across locations, no deep local knowledge | "I've seen this before. Am I seeing it again?" |
| 7 | Specialist | Valuable skill that grants unique access | peer in professional spaces, outsider socially | Deep technical knowledge, limited social knowledge | "They need me. Can I use that without being used?" |
| 8 | Local | Born here. Never left. | deep insider socially, peer/insider with grey economy through proximity | Knows everything about the place, nothing about the wider world | "This is my home. What am I willing to do to protect it?" |
Archetype transition mechanics
Transitions happen through player action + time + community response. The archetype isn't assigned — it EMERGES from the player's relationship with the community.
NEWCOMER ──(get job)──→ INSIDER ──(take responsibility)──→ OPERATOR
│ │
│──(accept mission)──→ INVESTIGATOR ──(rise)──→ AUTHORITY
│
│──(stay transient)──→ DRIFTER
│
│──(apply skill)──→ SPECIALIST
LOCAL ──(get drawn in)──→ INSIDER ──(take over)──→ OPERATOR
│
│──(leave home)──→ DRIFTER
INVESTIGATOR ──(drop credentials)──→ NEWCOMER (or DRIFTER)
AUTHORITY ──(lose position)──→ DRIFTER (or NEWCOMER)
OPERATOR ──(get caught)──→ prisoner/exile mechanic (future)
INSIDER ──(withdraw)──→ LOCAL (if they've been here long enough)
For generators: The archetype defines:
- Which NPCs are eligible to become this PC (e.g., Insider requires
faction_access: [ring], Newcomer requirestime_in_district: 0) - What monologue pool this archetype uses
- What the investigation/engagement style is
- What tools/approaches are available
- How NPCs default-react to this archetype
Cultural variation in archetype viability
Not every archetype is equally viable in every culture:
| Society Parameter | Archetype Effect |
|---|---|
authority.attitude: hostile |
Investigator and Authority start at a disadvantage; Insider and Local have stronger cover |
authority.attitude: absent |
Authority archetype unavailable; Investigator must work without institutional backing |
social.privacy_level: high |
Newcomer and Drifter face steep trust curves; Insider and Local have deep cover |
social.privacy_level: low |
Newcomer integrates faster; Insider's secrets are harder to keep |
solidarity_type: family |
Local has massive advantage (family connections); Newcomer is locked out without marriage/adoption |
solidarity_type: labor-informal |
Specialist gains quick peer access through demonstrated skill |
drift_stage: pioneer |
Everyone is a Newcomer; archetype distinctions are flatter |
economic_function: transit |
Drifter is the norm; Local barely exists |
This means: the SAME archetype plays differently in different cultures. An Investigator in a Frost/Salt culture (Krenn-like) faces patient silence. An Investigator in a Tide/Vine culture faces elaborate social performance. An Investigator in an Iron culture faces organized collective resistance. Eight archetypes times the ingredient variation space produces hundreds of distinct play experiences.
5. Sacred / Profane / Middle Kingdom — Worldbuilding Elements
Sacred (rules that never change)
| Element | Why Sacred |
|---|---|
| The ingredients menu exists (heritage roots, motivations, pressures, etc.) | The generation system's vocabulary |
| 10 heritage roots with defined parameter spaces | The phonetic/social flavor palette |
| 5 access tiers (public/peer/insider/authority/hostile) | The investigation mechanics foundation |
| Trust is earned, not given | Core gameplay principle |
| Every society has a grey economy | Core investigation premise |
| The FRIEND pattern (warmth → trust → doubt → conflict) | Emotional architecture |
| Gate network connects worlds | Physical infrastructure |
| Core worlds are always connected | Navigational stability |
| 9 thematic patterns + 6 functional motivations exist | NPC composition vocabulary |
| 8 archetype positions exist with transition paths | Player identity vocabulary |
Profane (parameters drawn from pools per playthrough)
| Element | Pool | Scale |
|---|---|---|
| Which heritage roots a specific world has | 10 roots × weights | Per world |
| Which economic pressure drives the grey economy | 7 pressure types | Per world |
| Which contraband type is being moved | Contraband pool per pressure type | Per district |
| Which NPCs fill which pattern/motivation slots | NPC generation pool | Per district |
| How long trust takes to build (culturally variable) | trust.building_rate parameter range |
Per society |
| What the authority attitude is | 5 attitude types | Per society |
| Where peripheral gates connect | Peripheral world pool | Per playthrough |
| Who the FRIEND is | FRIEND-eligible NPC pool | Per playthrough |
| Which worlds are accessible | Peripheral connection pool | Per playthrough |
| What economic function a specific world has | 10 function types | Per world |
Middle Kingdom (sacred structure, profane cast)
| Element | Sacred Part | Profane Part |
|---|---|---|
| Gate topology | Core hubs always connected. Minimum connectivity rules. | Which peripheral worlds connect where. Number of accessible systems fixed but which ones varies. |
| NPC population | 30/50/20 split (D-029). Pattern distribution constraints. | Which NPCs are which. Which NOBODYs have hidden depths. |
| Society profile | The parameter categories exist. | Which values are assigned per world. |
| Triangle structure | 2 per social site minimum. Triangles have 3 nodes with conflicting interests. | Which NPCs form which triangles. What the conflict is about. |
| FRIEND arc | 5 phases. Must have observable contradiction. Must exercise tell system. | Which NPC. Which contradiction. Which tell. |
| Investigation | Detective faces cultural resistance. Smuggler has embedded advantage. | How much resistance. What kind. How the advantage manifests. |
| Archetype transitions | Transition paths exist. Actions determine position. | Which transitions are available in this culture. How long they take. |
Gate topology as macro randomization
The gate network is the cultural map's skeleton. At 300 worlds:
Sacred structure:
- 3-5 core hub worlds are ALWAYS connected to each other (the Administrative Heart, the major trade nexuses). These anchor the game's navigational framework.
- Every world has at least 1 gate connection (no completely isolated worlds — isolation is social, not physical).
- Gate connections are bidirectional.
- Connection distance affects travel time and trade cost.
Profane variation:
- Peripheral worlds (200+ of the 300) have randomized connections per playthrough.
- Some peripheral connections create strategic chokepoints (a Frontier system that's the only route to a Research Outpost).
- The number of accessible worlds may be fixed (e.g., you can reach 40-60 worlds from any starting position) but WHICH ones varies per seed.
- Unlockable gates (via random gate events or quests) expand the accessible map during play.
Worldbuilding implications of gate topology:
- Worlds that are gate-adjacent develop TRADE RELATIONSHIPS and cultural blending. Two worlds connected by a direct gate share more heritage drift than two worlds separated by 3 hops.
- Chokepoint worlds develop TRANSIT CULTURES (the Gate Junction archetype from my Round 3). These are Middle Kingdom — the transit function is sacred, the specific population is profane.
- Isolated peripheral worlds develop MORE EXTREME cultures (less blending, more drift, stronger heritage dominance). The generator should increase
drift_modifiers.novelfor isolated worlds. - The player's starting position determines which cultural contexts they encounter. A playthrough starting in the Frost/Salt cluster sees different worlds than one starting in the Tide/Vine cluster. This is the DEEPEST replayability axis — the cultural landscape itself shifts.
Gravity and perception (flagging for perception team)
The lead flagged varying gravity as a perception element. From a worldbuilding perspective:
- Station worlds = standard gravity (simulated). Feels like home for long-term residents.
- Surface worlds = variable gravity. Krenn Prime has a specific gravity that station-born visitors notice.
- Low-gravity stations = social markers. People move differently, carry differently, fight differently.
- Gravity variation is a WORLDBUILDING INGREDIENT, not just a physics parameter. A society that lives in low gravity for 200 years develops different body language, different spatial norms, different architecture. The heritage root blending should include a
gravity_normparameter.
# Addition to society profile
environment:
gravity_norm: 0.85g # affects NPC movement, architecture, body language
gravity_variation_notable: true # do locals notice gravity? (station-born notice planetary surface)
6. IP Originality at 300-World Scale
Setting note — this is where originality risk is HIGHEST.
At 10 regions, I could manually verify each culture against known SF properties. At 300 worlds generated from an ingredients menu, the generator will inevitably produce combinations that FEEL like specific franchises:
| Risky combination | Feels like | Mitigation |
|---|---|---|
| Arc heritage + academic settlement + meritocratic authority | Banks's Culture (post-scarcity intellectual society) | The Culture lacks the heritage-drift mechanic and the grey economy. Our version has economic pressure and smuggling. Distinct. |
| Iron heritage + extractive economics + debt-trap pressure | Alien franchise (company-colony, exploited workers) | Common SF trope, not specific enough to one franchise. Our version has the ingredients system rather than a monolithic "evil corporation." |
| Frost heritage + logistics + tight-margin + prohibition | Hamilton's Commonwealth (only if we use Commonwealth-specific names/tech) | Already verified in Rounds 1-2. Krenn is structurally original. |
| Dust heritage + communal + survival-gap | Le Guin's Anarres (The Dispossessed) | Our version has a grey economy; Anarres explicitly doesn't. Distinct. |
| Jade heritage + hierarchical + institutional-military | Generic "space samurai" trope | Heritage drift should prevent this. At mature drift, the Jade root is substrate, not surface. The society shouldn't feel like "Japan in space." |
Mitigation strategy: The ingredients menu includes a cultural distance validator — a generator constraint that checks whether a produced society falls within a known-franchise similarity threshold. If the combination is too close to a known property (measured by surface-feature overlap), the generator adjusts one ingredient.
This is a generator-level constraint, not a human review step. At 300 worlds, we can't manually review every culture for franchise similarity. The validator must be automatic.
The deepest defense against franchise similarity is the DRIFT MECHANIC. A "mature" or "ancient" drift stage blends heritage roots so heavily that the surface features of any single Earth culture are dissolved. A Jade/Stone society at 300 years of drift doesn't look Japanese — it looks like whatever emerged from precise-phonetic + land-connected social norms + 300 years of station life. The DRIFT is what makes our cultures original.
Summary of Round 4 Deliverables
| Deliverable | Status |
|---|---|
| Ingredients menu (10 heritage roots + 5 other categories) | Defined above with generator parameters |
| Mechanical parameter matrix (society profile spec) | Full YAML spec above with per-system outputs |
| NPC pattern × motivation composition rules | Rules + distribution constraints + cultural modifiers |
| 8 PC archetypes | Defined with access maps, moral positions, transitions |
| Sacred/Profane/Middle Kingdom mapping | All worldbuilding elements classified |
| Gate topology randomization rules | Core/peripheral structure defined |
| IP originality strategy at scale | Drift mechanic as primary defense + automated similarity validator |
What changed from Round 3
| Round 3 (20 districts) | Round 4 (300 worlds) |
|---|---|
| 10 fixed conceptual regions | 10 heritage roots that combine into thousands of societies |
| Regional brief as documentation | Society profile as generator output |
| Hand-authored naming pools | Algorithmic naming from phonetic parameters |
| Prose social norms | Numeric parameters (privacy_level: 0.8) |
| Each region needs a worldbuilder | Worldbuilder designs the ingredient menu; generators produce cultures |
| Quality from expert review | Quality from constraint specification |
What I need from other agents
- Tyre: Can the content pipeline consume this parameter matrix format? Is
serde_yamlsufficient for this schema depth? - Gestalt: Do the pattern × motivation composition rules create enough mechanical variety? Are the distribution constraints too tight or too loose?
- Nigel: Does the Sacred/Profane/Middle Kingdom mapping protect replayability at 300-world scale? Are there randomization axes I missed?
- Gore: Does the economic pressure ingredient capture the thematic weight? Does "tight margin" vs. "survival gap" vs. "prohibition economy" create meaningfully different moral textures?
- Mellanie: Can the
environment.*parameters in the society profile give you enough to generate environmental text? What's missing?
Status: Round 4 response complete. Author: Miri Date: 2026-02-12