workflow-optimization

Workflow Optimization

A workflow definition has accumulated unnecessary steps over time. Some tasks run sequentially when they could run in parallel; others fetch data that is never used downstream. The optimization pipeline analyzes the workflow DAG, identifies parallelization opportunities, detects dead tasks, and produces an optimized definition.

Pipeline

[wfo_analyze_execution]
     |
     v
[wfo_identify_waste]
     |
     v
[wfo_parallelize]
     |
     v
[wfo_benchmark]

Workflow inputs: workflowName, executionHistory

Workers

WfoAnalyzeExecutionWorker (task: wfo_analyze_execution)

• Writes dependencyGraph, taskTimings, originalOrder, totalDurationMs

WfoBenchmarkWorker (task: wfo_benchmark)

• Reads originalDurationMs. Writes benchmarkResult

WfoIdentifyWasteWorker (task: wfo_identify_waste)

• Writes wasteItems, independentTasks, totalWasteMs

WfoParallelizeWorker (task: wfo_parallelize)

• Writes optimizedPlan

---

16 tests | Workflow: wfo_workflow_optimization | Timeout: 60s

See RUNNING.md for setup and usage.