mpi-execute-parallel

To run parallel tasks safely, an orchestrator session coordinates worker sub-agents, validates ownership rules, and applies safety checks to prevent conflicting edits during parallel batch execution.

A parallel batch is an explicit section in a large plan designated for simultaneous execution. Worker sub-agents claim ownership of these tasks to maximize throughput while the orchestrator reports task-board changes.

To start a parallel batch, request execution using prompts like 'MPI execute parallel' on an active plan containing a Parallel Batch section. The orchestrator then guides worker assignment and plan updates automatically.

Yes, coordinating parallel worker agents requires access to the mpi-lib shared references. This library provides the foundation for the orchestrator session to manage task ownership and synchronize the task-board.

Parallel task execution halts during ownership validation failures because the orchestrator enforces strict safety checks to prevent conflicting edits. This mechanism ensures plan and task-board integrity when multiple workers attempt to claim the same task.