research-harness-plan

To generate an executable analysis plan from a study specification, you need to synthesize study_spec.md, data_audit_report.md, and data_cleaning_report.md into a concrete, task-level analysis_plan.md with defined inputs, outputs, and definition of done.

A reproducible research workflow plan ensures traceable, audit-ready task definitions before analysis scripts run. It validates study variables against cleaned data, flags missing variables as unknown, and prevents impossible analysis tasks.

Validating which study variables exist in cleaned data requires feasibility and data checks that cross-reference the approved study specification with the data cleaning report, automatically flagging any missing variables as unknown to avoid impossible tasks.

Yes, automated plan synthesis can produce a ready-to-execute analysis plan covering descriptive statistics, main models, robustness checks, and figures by mapping approved study specifications and audit findings to concrete task-level scripts and outputs.

Defining the definition of done for empirical analysis tasks involves generating an analysis_plan.md that specifies complete task lists, required scripts, expected logs, defined outputs, and pre-flight checks for each analysis stage.

If the approved study specification includes variables missing from the cleaned data structure, the plan generator flags those missing variables as unknown during feasibility checks to prevent generating impossible analysis tasks.