Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.wazoo.dev/llms.txt

Use this file to discover all available pages before exploring further.

Autonomous agents write new facts into the graph to evolve the world. Worlds supports multiple strategies for ingesting and synchronizing knowledge within a stateful context.

​
State mutations

A world mutates its state using RDF patches, which are granular additions and deletions of facts. Because Worlds functions as a chronological, append-only ledger, these patches permanently preserve historical truth.

​
SPARQL updates

The most common way to mutate state is executing SPARQL updates. When an agent runs an INSERT or DELETE command, Worlds translates it into a deterministic patch.
SPARQL
PREFIX wazoo: <https://wazoo.dev/#>
PREFIX schema: <https://schema.org/>

INSERT DATA {
  wazoo:organization a schema:Organization .
}
Changes in Worlds are orchestrated through the API or direct SPARQL updates. Unlike traditional databases that overwrite records, Worlds utilizes RDF patches to maintain a chronological system of record.

​
Mutation and state

Every update to a world is a transaction that appends new facts to the ledger. This ensures that agents can always query past states and understand how information has evolved over time.

​
RDF patches

An RDF patch is a structured set of additions and deletions. When you update a relationship, the engine performs a patch:
  1. Deletion: Remove the outdated triple.
  2. Addition: Insert the new, verified triple.
This process is atomic and verifiable.

​
Update strategies

Worlds supports multiple methods for mutating state, depending on the required precision and automation level.
MethodPrecisionUse case
SPARQL UpdateHighestComplex, conditional logic and batch deletes
API assertHighProgrammatic ingestion of verified facts
RDF PatchesHighDelta-based synchronization from external stores

​
Feedback ingestion

Intentional agency requires a bridge between human preferences and graph state. RLHF enables this by treating feedback as a first-class mutation.

​
Capturing rewards

When a user or supervisor provides feedback (e.g., a thumbs up or a specific correction), the API records this as a preference item connected to the original fact.
TypeScript
import { Worlds } from "@wazoo/worlds-sdk";

const worlds = new Worlds({
  apiKey: process.env.WORLDS_API_KEY,
});

// Execute a SPARQL update to assert a preference (reward)
await worlds.sparql(
  "my-world-id",
  `PREFIX worlds: <https://schema.wazoo.dev#>
  PREFIX user: <https://etok.me/#>

  INSERT DATA {
    <fact:123> worlds:hasPreference [
      a worlds:Preference ;
      worlds:hasReward 1.0 ;
      worlds:verifiedBy user:person
    ] .
  }`,
);

​
Recursive learning

As preferences accumulate, the world’s probability landscape is reshaped. Retrieval results are boosted based on historical reward signals, creating a recursive loop where the system learns which triples are most useful or truthful for the agent’s specific context. This historical truth remains in the ledger, allowing future queries to be guided by this preference data. Learn more about steered retrieval in the Alignment deep-dive.