run-manager

Run Manager orchestrates FABER workflow runs by generating unique run IDs, initializing run directories and state, and emitting structured events through a central gateway. This enables you to coordinate parallel executions, resume from failures, and maintain a complete audit trail across all run phases.

Run_id validation enforces unique identifiers before operations, preventing duplicate runs and ensuring each workflow execution is distinctly tracked. Combined with atomic state updates and event emission, this prevents corruption and maintains consistency across parallel or resumed runs.

Run Manager supports resuming from failures and rerunning with modified parameters by maintaining isolated run directories and state for each run_id. Your workflow can recover mid-execution or restart with new settings while preserving the audit trail from prior attempts.

Yes. Run Manager uses bash scripts and jq for JSON processing to generate run IDs, manage state, and emit events. This dependency stack lets you integrate workflow orchestration into shell-based automation pipelines without additional runtime requirements.

Run Manager prevents direct event modification, deletions without archival, and path-traversal attempts through run_id validation and atomic writes. These constraints protect the integrity and auditability of your workflow event log across all run phases.

Run Manager emits structured events throughout the workflow lifecycle, creating a complete audit trail. This event log enables you to trace execution history, debug failures, and meet compliance requirements by recording every state transition and operation.