Meridians is a causal simulator for decisions under consequence and training. Bring a consequential situation; the system turns it into a kept world, runs the world forward, lets you act inside it, and preserves what happened so you can ask: what did I cause?
Meridians Wiki
The maintained map of Meridians
Explore the concepts, records, practices, and specifications that make Meridians a causal simulator for decisions under consequence and training.
Signed out — public pages only
Categories
A–Z indexFoundations12 articles
The principles and premises Meridians is built on.
Product58 articles
What Meridians does and where it is going.
Vocabulary83 articles
The terms that make the system legible.
Architecture22 articles
The system's technical shape and boundaries.
Operations6 articles
How the system is run.
Conventions24 articles
Shared working agreements.
Worlds4 articles
World material and developed settings.
Plans1 articles
Plans and directional work.
Specifications20 articles
Stable contracts and technical specifications.
Personas14 articles
People and roles the system serves.
Insights1 articles
Learnings and analysis.
Diagrams52 articles
Visual explanations and diagrams.
A–Z index
3
A
A smaller interaction grammar[Public]Access gate [Flow][Public]Acting and Entity Interpretation — evidence-grounded actor context[Public]Adding work — the rule[Public]AI model routing — tier · category · profile [Concept][Public]AI models[Public]AI Models — tiers, categories, and allocation profiles[Public]Akashyatra — founding world contract[Public]Akashyatra canon source — Chen and the Seeds[Public]Akashyatra canon source — Taylor and Crystallisation[Public]Alignment — legible worlds and accountable agency[Public]Ana Duarte — Explorer[Public]Anti-personas — who Meridians is not for[Public]API & State Principles — the grounded vocabulary and handling rules[Public]Architecture[Public]Architecture — runtime boundaries, canonical state, and the one-writer model[Public]Architecture overview[Public]Architecture tree[Public]Article[Public]Auth & access surfaces [UI][Public]Authorship and provenance — who made what[Public]Autopilot / Pilot[Public]
B
Beat Classes[Public]Behaviour flows[Public]Belief and verification[Public]Belief System[Public]Belief vocabulary[Public]Book[Public]Branch refinement — Meridians World Search (MWS): Monte Carlo tree search over a recorded world[Public]Branching — the fork instrument and the shape of the novel[Public]Brand copy[Public]Budgeted fate exploration — scouting timelines worth producing[Public]Bundle.[Public]
C
Cache coherence — server-of-record [Flow][Public]Calibration[Public]Calibration[Public]Canon channel[Public]Categories[Public]category-positioning[Public]Center views [UI][Public]7. Change protocol[Public]confidence at entry[Public]Consequence[Public]Constellation role / domain override.[Public]Constellation surface [UI][Public]control-surface[Public]Conventions knowledge[Public]Coordination presets[Public]Copy and localisation[Public]Coverage[Public]6. Current gaps to close[Public]
D
Decision architecture — building up to the fork[Public]Dev Ramanathan — Operator[Public]Directive lifecycle [Flow][Public]Director-governed onboarding (MCP).[Public]discipline[Public]3. Documents — daemon-first, then echoed to the cache[Public]14. Domain categories — src/lib/client/store/store.tsx[Public]Domain creation [Flow][Public]Domain Position[Public]Domain settings [UI][Public]Domain spine — System · World · Fate [Concept][Public]dormant[Public]
E
Editorial[Public]Elena Marquez — Author[Public]Engagement systems — decision weight, accumulation, and what the medium already solved[Public]Engine pipeline — data · AI · persistence [Flow][Public]Entity Interpretation Graph — evidence into selected context [Concept][Public]Exchange.[Public]Expert paradigm corpus (Onboarding).[Public]Extraction jobs [UI][Public]Extraction pipeline — text → domain [Flow][Public]
F
G
H
headline[Public]Hill-climbing opportunities[Public]Home console [UI][Public]7. Hosted exposure — the one open liability[Public]Hosted sign-in across instance domains [Flow][Public]House voice and usage[Public]5. How the restructure was done (and the one thing left)[Public]Human loop — Orient → Judge → Steer → Learn → Refine.[Public]Human loop [Flow][Public]Humanistic priors[Public]
I
K
L
Landing-page sequence[Public]Latent traits — the practice record as an instrument that steers generation[Public]layers[Public]learning[Public]Legacy lifecycle — how compatibility enters, rolls through the fleet, and leaves[Public]Leo — Builder[Public]listing[Public]LLM context hydration[Public]LLM wiki operations[Public]Local insight bundles[Public]
M
Maintenance creation + the research run [Flow][Public]Maya Chen — Facilitator[Public]MCP action chains — how an AI navigates the surface [Flow][Public]MCP runtime — transports, trust, and attribution [Flow][Public]5. Measurement[Public]Measures[Public]memory-and-context[Public]Merge[Public]Merge resolution — Streams + Notes → Merge → Graph → Arc [Flow][Public]Meridians[Public]Meridians documentation[Public]Meridians worlds[Public]Message library[Public]Mobile [UI][Public]monitor[Public]
N
O
Onboarding — build a Constellation from intent [Flow][Public]Onboarding / Onboarding document.[Public]Onboarding workspace [UI][Public]Operation lifecycle — a Tier-2 op, dispatch to commit [Flow][Public]Operations timezone[Public]19. Outbound research, monitoring & the Sources repo[Public]Outcomes are the universal frame.[Public]
P
Pentad archetypes[Public]Persistence — where state lives and how it propagates[Public]Persistence and sync[Public]Personas — who reaches for Meridians[Public]Phase graph — the working model creativity inherits [Concept][Public]Plans knowledge[Public]Platform model — two products, three authorities, one daemon [Concept][Public]platform-and-people[Public]play[Public]Position[Public]Position (as the graded unit)[Public]Position cycle[Public]Position lifecycle [Flow][Public]Positioning — product identity and boundary[Public]Positioning tree[Public]Preparation / extraction (Onboarding).[Public]Prior[Public]Priors[Public]Private Domain / Public Domain.[Public]Priya Nair — Reader[Public]Product behaviours[Public]Product behaviours — building and inhabiting graph-worlds[Public]Product economics[Public]Product knowledge[Public]Product rules[Public]Product strategy[Public]product-model[Public]Production — the asset inventory, the style lock, and what a novel costs to look like[Public]Profiles — the allocation matrix[Public]Program — Research → Opinion → Merge → Position.[Public]Projection anatomy — Reading · Graph · Signal [Concept][Public]
R
Rafael Ortiz — Strategist[Public]re-prices the read[Public]Reading[Public]reasoning[Public]Reasoning graph engine — three entry points [Flow][Public]Reducer taxonomy — one writer, many slices [Flow][Public]Release & deploy [Flow][Public]Reminder.[Public]Renata Sokolova — Table[Public]11. Replication model[Public]repositions[Public]Research pipeline — sources → Firecrawl → the Research pass [Flow][Public]Right inspector [UI][Public]Risks (strategic register)[Public]8. Risks & open questions[Public]Run · Directive · Event (the Program machinery)[Public]
S
Sam Okafor — Participant[Public]Scenario — the game loop [Flow][Public]Scenario UI — the poker table [UI][Public]Scene cadence (Pentad signs)[Public]Schedule · Cadence (how cycles + Positions compose into Research)[Public]Scheduler — cadence to fired run [Flow][Public]1. Services & real 2026 costs (mapped to env vars)[Public]settle into canon[Public]Signal[Public]Signals[Public]snapshot[Public]Sonic — micro analysis of prose as heard composition[Public]Source ledger[Public]Source plan (Onboarding).[Public]2. Sources of truth — the exact words[Public]Specs — how the bounded proposals compose[Public]Stance[Public]Static-analysis maintenance[Public]Story / Stories[Public]Story pipeline — events → a legible feed [Flow][Public]Story Scales[Public]strategic-play[Public]Strategy — where we are going and what we will not claim[Public]Stream[Public]Stream lifecycle [Flow][Public]Streams[Public]STRUCTURE — how the codebase is organized (agent-first)[Public]substrate[Public]3. Surface map[Public]Sync protocol — master ⇄ clients [Flow][Public]System Connections — runtime and surface wiring map[Public]System diagrams[Public]
T
The 2026-09-01 pivot — Explorable VN worlds as the consumable artifact[Public]The behavioural loops[Public]The causal-simulator monopoly — what we own, why it is sci-fi, and the order we build it[Public]The emotional progression[Public]1. The funnel thesis[Public]The grove — 木 and Meridians World Search (MWS) on a story[Public]The human and agent experience[Public]4. The layer map — src/lib/ is role-first[Public]The market[Public]The Memory Palace — the experiential frame[Public]The moat, ordered by what survives stronger frontier models[Public]0. The model (run it)[Public]The north star[Public]4. The one-continuation rule[Public]The pentad — five independent forces, observed couplings, and a pipeline that can be reordered[Public]The pentad in the app — measurement, surfaces, post-hoc actions, roadmap[Public]The pitch hierarchy[Public]The practice record — the journey, the measure, the proof[Public]6. The pricing model — flat fee + usage-based above an allowance[Public]The product is a practice[Public]The Program loop [Concept][Public]The Reader — a dedicated visual-novel interface[Public]The same practice, headless — one loop, three drivers[Public]9. The simulation as the cost unit[Public]The spine[Public]The strategic logic (five load-bearing beliefs)[Public]4. The three cost layers[Public]The three layers[Public]The three opinion layers (memory · action · foresight).[Public]The two systems[Public]The VN scaffold — Script as the projection, summary and deltas as the truth[Public]3. The write path (follow one action end to end)[Public]Thesis[Public]Threads[Public]three complementary belief layers over different time horizons[Public]Three layers of opinion [Concept][Public]Tiers[Public]TL;DR[Public]Tomás Ferreira — Learner[Public]TopBar menus & run surfaces [UI][Public]Track record[Public]Training worlds — requirements for an hour that is remembered[Public]Training worlds for agents[Public]Trust weight[Public]Two magics, five magics[Public]
U
V
W
Weekly update — Friday GitHub business summary [Flow][Public]2. What each stage must sell[Public]What separates the three layers is the write path, not the kind of question they ask.[Public]What survives / promotes / recasts[Public]What we don't do[Public]Where it's controlled[Public]7. Where to read next[Public]5. Why a *pure* flat fee breaks (and what fixes it)[Public]Why Stories fits[Public]Wiki classification policy[Public]Workspace shell [UI][Public]World — cumulative vision corpus[Public]World — the Domain as a place you can wander[Public]World — the living graph-world[Public]World Density[Public]World engine direction — stable boundaries for branch-derived Worlds[Public]World Spirit[Public]World tree — the many-worlds narrative temporal tree, and the shared multiverse the VNMMO is made of[Public]
This Wiki is a transparent port of the company's internal knowledge base. Every article names its source path so you can follow the maintained record back to its owning document.
What it is
Meridians keeps a single canonical record of a world: its actors, incentives, private information, rules, constraints, evidence, and open questions. It is not a prediction oracle. A person, seat, Director, or agent supplies Will from outside the simulation; Meridians computes and records consequences.
How a Domain works
A Domain is the kept world. It holds World, System, and Fate as stored forces, while Theme and Causality are derived readings. Threads carry the long horizon; Streams carry present-day priors; Positions hold commitments grounded in a Reasoning Graph. Scenes are the unit of change, and branches preserve parallel histories rather than replacing one another. Start with the Domain model and the architecture.
Deciding in Scenario
Scenario is where a person meets a branch tip. Choose a move, attribute it to its source, and let the world settle the continuation. A Reading is a priced fan of futures from an as-of vantage, not a verdict. MWS is the specified search that will develop uncertain, costly forks and price pathways; it proposes and prices, never chooses.
What it does not do
Meridians does not claim to predict the real world, verify itself against markets, or turn scores into verdicts about people. Fictional and synthetic results describe the configured world unless separately validated against external evidence. Voice intake, automated MWS, latent-trait readings, and a full VN reader remain direction rather than universally shipped capability.