workflow-inception

Automated task selection and execution loops are coordinated by selecting backlog items by priority, verifying acceptance criteria, and committing changes per policy without manual intervention. The workflow-inception orchestrator executes this loop deterministically, enforcing max_tasks limits and generating run logs for each iteration.

An automated inception loop is a process that aggregates configuration, bootstrap, and git state to select, perform, verify, and finalize development tasks iteratively. It removes manual coordination by enforcing deterministic task selection, preflight checks, and run finalization within Volon-managed repositories.

Bootstrapping new work and promoting backlog items is handled by selecting the next work unit via bootstrap hint or priority, updating task status, and regenerating bootstrap state after finalization. This ensures continuous iteration while preserving commit and logging policies.

The automated task orchestrator supports both iteration-or-task commit modes, allowing changes to be committed either per individual task or per entire run. It enforces commit policies alongside max_tasks limits and writes factual run logs to track progress.

Headless task automation can complete up to N high-priority backlog items by running the inception loop, verifying acceptance criteria, and committing changes according to policy. It operates without external model invocation, relying on deterministic selection and bootstrap regeneration.

Deterministic task orchestrators are limited to Volon-managed repositories and rely on preflight checks, bootstrap hints, and priority queues for task selection. They do not invoke external models, requiring all configuration, backlog state, and acceptance criteria to be defined beforehand.