hypervisor

What problem does it solve?

Coordinate batch operations with atomic locking to prevent duplicate processing of tasks across multiple workers.

Core Features & Use Cases

• Atomic locking: Use mkdir to atomically claim a task directory and prevent other workers from processing the same item.
• Queue-driven processing: Maintain a queue of work items and a results directory to track progress and outcomes.
• Parallel workers: Support multiple agents running in parallel while ensuring each item is processed once.
• Result tracking & triage: Write per-item YAML results and provide a lightweight triage workflow to classify or close tasks.

Quick Start

Run batch_worker.py with a prepared queue at /tmp/task-batch to start processing items in parallel using atomic directory locks.