ralph-loop

An autonomous loop continuously refines a task by orchestrating iterative work cycles until completion. You start it by running the loop command with your task description, and it manages the iteration flow automatically.

State persistence in autonomous loops saves progress to a local state file, allowing iterative task management to resume seamlessly across sessions. The loop stores its current status at a local path to maintain context.

To prevent an autonomous loop from running indefinitely, you configure max_iterations and completion_promise parameters. These settings guard against runaway runs by setting hard limits on cycles and defining completion criteria.

The stop-hook integration manages safe exit during iterative task cycles by intercepting the loop flow between iterations. It ensures the autonomous process halts gracefully when completion triggers or iteration limits are reached.

Yes, autonomous looping supports cross-session task management without manual prompting. It uses a persistent state file to track progress, automatically continuing the task refinement cycle across interruptions until the defined completion promise is met.