Agimus Python SDK

Query and edit your ontology from Python. Discover its schema, filter and traverse objects, compute aggregates, execute read-only SQL, and retrieve typed Arrow tables. Synchronous and asynchronous clients share the same query builder.

Install

pip install agimus

Requires Python 3.10+. Install pandas or polars separately for their DataFrame conversions.

Connect and read

Create a key in your tenant's API Access page. Keep it outside source code:

export AGIMUS_API_KEY="agm_..."
# Optional; the default is https://api.agimus.ai
export AGIMUS_BASE_URL="https://api.agimus.ai"

from agimus import AgimusClient

with AgimusClient() as client:
    print([entity["apiName"] for entity in client.list_entities()])

    # Replace these API names with those in your ontology.
    customers = client.objects("Customer")
    page = customers.filter(region="EU").fields("customerId", "name").page()
    for customer in page:
        print(customer)

For writes, use a read/write key and retain the version from your read:

with AgimusClient() as client:
    customers = client.objects("Customer")
    customer = customers.get("C1")
    updated = customers.update("C1", {"name": "Acme Ltd"}, version=customer["_version"])

The SDK reads and writes ontology objects. Permissions apply to discovery, queries and writes. Large result delivery is followed automatically; resumable handles let you continue an execution after the client's deadline.

Documentation

Sign in to the Agimus app and open Resources → SDK & API Docs for the SDK guide, SDK reference and HTTP API reference. These cover discovery, reads, relationships, aggregates, SQL, Arrow, writes, async usage, parameters, result types, errors and limits.

License

MIT