orchestrate

DAG execution orchestrates multi-task plans by parsing layered dependencies and executing tasks sequentially or in parallel. It coordinates workflow execution automatically, applying safety checks and optional judge reviews to ensure reliable task completion.

Parallel task execution processes independent tasks simultaneously across DAG layers, while dependency-based failure handling ensures that downstream tasks are safely skipped or halted if their upstream dependencies fail, preventing cascading errors.

Automated execution reporting generates an execution_report.json file after completing DAG-based task plans. This report provides auditing and review details by capturing the execution status, safety check results, and task outcomes across all layers.

Yes, DAG-based workflow orchestration uses worktree isolation to separate per-layer parallel execution environments. This ensures that concurrent tasks operate independently without interfering with each other's files or execution state.

Yes, you can validate task plans before execution. The orchestration workflow applies validation and safety checks to the parsed DAG structure, ensuring that the sequential or parallel task layers are safe to execute.

Use DAG-driven orchestration when your multi-task plans contain complex layered dependencies requiring parallel execution, safety checks, or judge reviews. Simple sequential execution lacks the dependency-based failure handling and worktree isolation needed for reliable parallelism.