decomposition-scorer

Task decomposition scoring identifies and quantifies conflicts in planned task decompositions to ensure safe parallel execution. It evaluates file overlap, shared state, import chains, test coupling, and schema drift to determine isolation quality.

You can detect file overlap and shared state conflicts by running a decomposition audit through a four-phase workflow: gather the task list, build a conflict matrix, score isolation quality, and generate a structured risk report.

Score task isolation quality by evaluating conflict dimensions like file overlap, import chains, test coupling, and schema drift. The scorer applies a four-phase workflow to generate a quantified isolation score and execution order.

Yes, you can get execution order recommendations. The decomposition scorer generates a report that provides a recommended execution order and re-decomposition guidance based on the quantified conflict matrix.

To ensure safe parallelism in a codebase, check conflict dimensions including file overlap, shared state, import chains, test coupling, and schema drift. These dimensions determine the isolation quality of your task decomposition.

You should re-decompose tasks when the decomposition scorer identifies high conflict levels in file overlap, shared state, or import chains. The generated report provides actionable recommendations to restructure tasks for safer concurrent execution.