/api/workflowsList workflows
Returns templates first, including the Private Market Diligence Review reviewer path.
Workflow design, local demo execution, approvals, event normalization, and audit output.
https://lyzr-d-production.up.railway.appThe local proof-of-work does not require auth. A production deployment should add auth, tenant isolation, and stricter CORS settings before exposing workflow execution.
SyncFlow keeps the existing execute endpoint stable. Local mode extends the response with frontend-ready events and output, while Temporal mode keeps the original async shape and relies on WebSocket events for live state updates.
/api/workflowsReturns templates first, including the Private Market Diligence Review reviewer path.
/api/workflowsCreates a blank workflow definition with nodes, edges, and a generated operator-friendly name.
/api/workflows/:id/executeStarts local or Temporal execution depending on EXECUTION_BACKEND.
/api/approvals/:executionId/approveRecords approve or reject, then resumes local execution or signals Temporal.
curl -X POST "https://lyzr-d-production.up.railway.app/api/workflows/template-private-market-diligence/execute" \
-H "Content-Type: application/json" \
-d '{
"input_data": {
"input_text": "Revenue grew 42% YoY, gross retention is above 90%, and enterprise pipeline is concentrated in six accounts.",
"source_name": "sample-memo.txt"
}
}'Frontend-ready execution timeline in local mode.
Approval modal payload when the run pauses.
Final report data after a completed local run.
The execute endpoint can return events, output, and pending_approval immediately.
Temporal mode keeps the async contract and streams state over WebSockets.
Both paths use the same frontend helper before updating node status or approval state.
Approval responses can include continuation events and final output in local mode.
{
"execution_id": "...",
"workflow_id": "...",
"status": "waiting_approval",
"execution_backend": "local",
"events": [],
"pending_approval": {},
"output": null
}{
"status": "approved",
"execution_status": "completed",
"execution_backend": "local",
"events": [],
"output": {}
}