Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.
Python agentagentic-aiagentic-frameworkagentic-workflowaiautomationclaudedeepseekgenaigptllmlow-codemcpnextjsno-codeopenaipythonskillsworkflow
"""Unit tests for the RBAC bootstrap applied to a newly created agent."""
from collections.abc import Callable
import pytest
from services import rbac_agent_access_service
from tests.helpers.unit.rbac_fakes import Initialization, RBACDomain
def test_initialize_agent_rbac_access_seeds_scope_members_and_creator_policy(
rbac_domain: RBACDomain, config_overrides: Callable[..., None], monkeypatch: pytest.MonkeyPatch
) -> None:
config_overrides(RBAC_ENABLED=True)
def enqueue(tenant_id: str, account_id: str, **resources: str) -> None:
# The task must be queued before any remote writes.
assert not rbac_domain.transport.requests
rbac_domain.tasks.delay(tenant_id, account_id, **resources)
monkeypatch.setattr(rbac_agent_access_service.initialize_created_app_rbac_access_task, "delay", enqueue)
rbac_agent_access_service.initialize_agent_rbac_access(
tenant_id="tenant-1", agent_id="agent-1", creator_account_id="account-1"
)
assert rbac_domain.tasks.queued == [Initialization("tenant-1", "account-1", {"agent_id": "agent-1"})]
creator, whitelist = rbac_domain.transport.requests
assert creator.endpoint == "/rbac/access-policies/creator-member-bindings"
assert whitelist.endpoint == "/rbac/agents/whitelist"
for request in (creator, whitelist):
assert request.method == "PUT"
assert (request.tenant_id, request.account_id) == ("tenant-1", "account-1")
assert creator.params == {"resource_type": "agent", "agent_id": "agent-1"}
assert whitelist.params == {"agent_id": "agent-1"}
assert whitelist.json == {"automatic_include_workspace_members": True}