§ 00 · Simulator test mode

Simulate reality

Describe a business and the system invents its world — people, customers, suppliers, orders — then streams it in as packets (notes, emails, OCR'd invoices, receipts, voice notes) through the real pipeline. Each space is its own isolated graph. Watch it grow, evolve, and be understood.

generation
…

New space

a business to simulate
The first packets arrive once the world is built (a few seconds). Switch to Live to watch them land, or open the space's Graph as it grows.

Spaces

running worlds
No spaces yet — describe a business and press start.
§ 01 · Live

Real-time ingestion

Packets arrive from the receptors, are normalized and stored, then comprehended into the knowledge graph. Every panel below updates the moment a message lands.

packets stored
0
all time
this session
0
captured since load
deduped
0
identical re-dumps
graph updates
0
extractions this session

Ingestion pipeline

receptor → graph
receptor
normalize
blob
stash media
store
append · dedup
extract
genome → graph

Packet flow

waiting for input…

Event log

Modality mix

stored, by type
§ 02 · Graph

Knowledge graph

Entities and typed relations the cell has extracted from your messages. Hover a node to isolate its neighborhood; click one to read the source messages and the reasoning behind every edge.

Canvas

No graph yet.

Create a business in the Simulator — its packets are comprehended into entities and edges, and this canvas grows live.

Genome

the cell's DNA
§ 03 · Ontology

Engineered ontology

The schema this cell is held to — the write-time arbiter for every extraction. Subsumption, domain/range typing, lifecycles, and disjointness axioms are enforced here before anything reaches the graph.

Classes

is-a hierarchy · lifecycles

Relations

domain → range

Proposals

extensions the cell wants — admit via CLI
§ 04 · Understanding

Synthesized beliefs

What the cell has come to understand — macro-beliefs synthesized across many messages: risks, recurring patterns, and summaries, each revised as the evidence changes.

Beliefs

risks · patterns · summaries
§ 05 · Ask

Interrogate the graph

Ask in natural language. The cell retrieves the relevant messages and entities, answers only from that evidence, cites the exact sources, and rates its own confidence by how much corroborates the answer.

Ask a question about your graph — e.g. "what did I spend at the hardware store?", "what's open?", "who owes me?". It researches: semantically retrieves the relevant messages and entities, answers only from that evidence, cites the exact messages, and rates its confidence by how much corroborates it.
§ 06 · Packets

Perceived ledger

Every message the system has perceived, newest first. Click any packet to replay how the cell made sense of it — its reasoning, the entities it drew out, and how it connects to the past.

seqmodalitysourcesummaryhashperceived
§ 07 · System

Runtime status

Live state of the receptors feeding the system, the extraction and embedding backends, and the packet store.

Receptors

inputs

    Extraction

    backends

      Store

      persistence