Skip to main content

Cyberwave is in Private Beta.

Request early access to get access to the Cyberwave dashboard.

The SDK surface for triggering workflows and managing alerts on a twin. For building workflows and the node taxonomy, see the Workflows feature pages.

​
Workflows

List, trigger, and monitor workflows programmatically: 
cw = Cyberwave()

for wf in cw.workflows.list():
    print(f"{wf.name} ({wf.slug}) — {wf.status}")

# Trigger by UUID or unified slug
run = cw.workflows.trigger(
    "acme/workflows/pick-and-place",
    inputs={"target_position": [1.0, 2.0, 0.0], "speed": 0.5},
)

run.wait(timeout=60)
print(run.status, run.result)
Start from a Workflow object, or look up by slug: 
wf = cw.workflows.get("acme/workflows/pick-and-place")
run = wf.trigger(inputs={"speed": 1.0})
run.wait()

wf = cw.workflows.get_by_slug("acme/workflows/pick-and-place")
List and filter past runs: 
runs = cw.workflow_runs.list(workflow_id="workflow-uuid", status="error")
for r in runs:
    print(r.uuid, r.status, r.error)
Check if a workflow is currently running: 
if cw.workflows.is_running("workflow-uuid"):
    print("Workflow is currently executing")
is_running() returns True when any run has status running, waiting, or requested.

Workflows Overview

Automated sequences, cloud vs. edge execution, and the node taxonomy.

​
Alerts

Create, manage, and respond to alerts on a twin. Alerts notify operators that action is needed (e.g. a robot needs calibration or a sensor reading is out of range). 
twin = cw.twin(twin_id="your_twin_uuid")

alert = twin.alerts.create(
    name="Calibration needed",
    description="Joint 3 is drifting beyond tolerance",
    severity="warning",
    alert_type="calibration_needed",
)
SeverityDescription
infoInformational notice
warningRequires attention
errorSomething is wrong
criticalImmediate action needed
Manage the alert lifecycle: 
for a in twin.alerts.list(status="active"):
    print(a.name, a.severity, a.status)

alert.acknowledge()
alert.resolve()
alert.silence()
alert.update(severity="critical")
alert.delete()
Bypass backend deduplication and always create a new alert: 
alert = twin.alerts.create(
    name="Calibration needed",
    alert_type="calibration_needed",
    force=True,
)

Alerts (Edge & Drivers)

Alert categories and operator notification on the edge.