genpark-wukong-workflow-automator

GenPark Wukong Workflow Automator is a zero-dependency autonomous task orchestration and resilience engine implementing discrete Petri net reachability and deadlock validation, decorrelated exponential backoff Dead-Letter Queue (DLQ) routing, and Amdahl's Law parallel cluster throughput forecasting.

flowchart TD
    A[Workflow Places & Transitions] --> B[Petri Net State Machine Simulator]
    B --> C[Deadlock & Unbounded Token Detection]
    D[Task Failure Stream] --> E[Exponential Jitter Backoff Engine]
    E --> F[Retry Dispatcher / DLQ Escalation]
    G[Task Volume & Worker Cluster Size] --> H[Amdahl's Law Parallel Solver]
    H --> I[Throughput Tasks/Sec & Completion Time]

Features

• Petri Net Deadlock Validator: Simulates token transitions across bipartite graphs to ensure zero pipeline deadlocks.
• Decorrelated Backoff DLQ Router: Automatically calculates exponential delay intervals and routes exhausted jobs to dead-letter queues.
• Amdahl Throughput Forecaster: Models parallel cluster efficiency and tasks-per-second scaling bounds.

Quickstart

from client import GenParkWukongWorkflowAutomator

automator = GenParkWukongWorkflowAutomator()
res = automator.forecast_workflow_throughput_amdahl(10000, serial_fraction=0.10, worker_nodes=8)
print("Projected Throughput:", res["projected_throughput_tasks_per_sec"], "tasks/sec")

License

MIT