parallel-task-executor

Execute independent tasks in parallel and aggregate results using Anthropic's Parallelization pattern. Use when tasks can be decomposed into independent sections (file-level, module-level, test-level) or when evaluating multiple approaches through voting. Automatically manages dependency graphs, dynamic worker pools, and result synthesis for 2-10x speedup.