SDK reference

Install Python

git clone --recurse-submodules https://github.com/AnythingGraph/AnythingGraph.git
cd AnythingGraph/sdk/python
pip install -e .
python examples/supply_chain.py

from anythinggraph import AnythingGraphClient

client = AnythingGraphClient()
print(client.connector.health())

Install Node.js

cd AnythingGraph/sdk/nodejs
npm install
node examples/example.js

import { AnythingGraphClient } from 'anythinggraph-sdk';

const client = new AnythingGraphClient();
const health = await client.connector.health();

Install Go

cd AnythingGraph/sdk/golang
go build ./...
go run ./examples/supply-chain/

import (
    "context"
    "github.com/anythinggraph/anythinggraph-sdk/anythinggraph"
)

client := anythinggraph.NewClient(nil)
health, err := client.Connector.Health(context.Background())

Health

client.connector.health()                    # Python
await client.connector.health()              # Node.js
client.Connector.Health(ctx)                 # Go

Ingest

Inject records into an installed playbook. context.playbook_id scopes field mappings and relationship routing. Returns rows created, relationships created, and landing-zone count.

# Python
result = client.connector.ingest(
    {
        "playbook_id": "organizational-graph",
        "source_label": "hr-export",
        "entity_name": "employee",
        "field_mappings": [],
    },
    {
        "records": [{
            "full_name": "Sam Lee",
            "email": "sam@acme.example",
            "corporation_name": "Acme Corp",
        }]
    },
)
print(result["relationships_created"])

// Node.js
const result = await client.connector.ingest({
  playbook_id: 'organizational-graph',
  source_label: 'hr-export',
  entity_name: 'employee',
  field_mappings: [],
}, {
  records: [{
    full_name: 'Sam Lee',
    email: 'sam@acme.example',
    corporation_name: 'Acme Corp',
  }],
});
console.log(result.relationships_created);

List landing zone

Failed or ambiguous ingest records awaiting human review.

client.connector.list_landing_zone()         # Python
await client.connector.listLandingZone()     # Node.js
client.Connector.ListLandingZone(ctx)          # Go

Approve landing zone record

Apply corrections and train field mappings. Python / Node only.

await client.connector.approveLandingZone(12, {
  entity_name: 'employee',
  corrected_values: { full_name: 'Sam Lee' },
  field_mappings: [
    { from_field: 'name', to_field: 'full_name', entity_name: 'employee' },
  ],
});

Delete landing zone record

client.connector.delete_landing_zone(12)     # Python
await client.connector.deleteLandingZone(12) # Node.js
client.Connector.DeleteLandingZone(ctx, 12)  # Go

Health

client.rdf_cache.health()

Load cache

Load Turtle into the active cache capsule. For playbook-scoped capsules, use the REST API with playbook_id.

client.rdf_cache.load_cache(turtle_text, replace=True)   # Python
await client.rdfCache.loadCache(turtleText, true)        # Node.js

Get cached Turtle

client.rdf_cache.get_cache()
await client.rdfCache.getCache()

Cache metadata

Returns version, turtle_bytes, updated_at_ms.

client.rdf_cache.cache_meta()
await client.rdfCache.cacheMeta()

Clear cache

client.rdf_cache.clear_cache()
await client.rdfCache.clearCache()

SPARQL query

SELECT on the active cache. For governed agents with per-playbook capsules, use the REST API with playbook_id and subject_id.

query = "SELECT ?row ?name WHERE { ?row a <...#Employee> ; <...#fullName> ?name }"
client.rdf_cache.sparql_query(query)
await client.rdfCache.sparqlQuery(query)

Sync from data layer

Export Turtle from the data layer and load into the cache in one call.

client.rdf_cache.sync_from_data_layer(entity_ids="*")           # Python
await client.rdfCache.syncFromDataLayer({ entityIds: '*' })     # Node.js

Health

client.data_layer.health()

Entity CRUD

Manage record types (schemas) and fields.

entities = client.data_layer.list_entities()
entity = client.data_layer.get_entity(1)
client.data_layer.create_entity({
    "name": "employee",
    "display_name": "Employee",
    "fields": [
        { "field_name": "email", "field_type": "TEXT", "is_required": True }
    ],
})

Entity rows

Optional RebacOptions on listEntityRows.

rows = client.data_layer.list_entity_rows(
    entity_id,
    rebac=RebacOptions(playbook_id="organizational-graph"),
)
client.data_layer.create_entity_row(entity_id, {"email": "sam@acme.example"})

Row relationships

Traverse graph links between rows. Pass playbook_id to scope reads.

# Python — list employed_by links for a playbook
links = client.data_layer.list_relationships(
    rebac=RebacOptions(playbook_id="organizational-graph")
)

# Node.js
const links = await client.dataLayer.listRelationships({
  playbookId: 'organizational-graph',
});

client.data_layer.create_relationship({
    "relationship_name": "employed_by",
    "subject_entity_id": 1,
    "object_entity_id": 2,
    "subject_row_id": 10,
    "object_row_id": 20,
})

Export Turtle

Returns text/turtle for loading into the Data Cache client.

turtle = client.data_layer.export_turtle(entity_ids="*")
turtle = client.data_layer.export_turtle(entity_names=["employee", "corporation"])

List playbooks

playbooks = client.dashboard.list_playbooks()   # Python
await client.dashboard.listPlaybooks()           # Node.js

Install playbook

Creates entities, schema relationships, and optional example data in the data layer.

await client.dashboard.installPlaybook('organizational-graph')

Playbook webhook ingest

Simplest ingest path — dashboard resolves mappings and forwards to the connector.

await client.dashboard.playbookWebhook('organizational-graph', {
  records: [{
    full_name: 'Sam Lee',
    email: 'sam@acme.example',
    corporation_name: 'Acme Corp',
  }],
}, 'hr-export');