hydromind-research-execute

The research workflow progresses through six canonical phases: prepare, implement, configure, run, analyze, and summarize. The execute skill orchestrates this sequence to advance hydrology paper-directed experiments while maintaining strict data and artifact ownership.

You validate controlled writes during hydrology research by enforcing the use of design and paper inputs alongside hydro CLI surfaces such as hydro config, hydro run, hydro queue, and hydro checkpoint to control and validate experiment artifacts.

The evidence-loop workflow manages and protects canonical outputs including execution/state.json, execution/loops/loop{n}.md, execution/results.md, and execution/evidence_review.json to ensure reproducible artifact progression.

Yes, design and paper artifacts are required. The system loads required artifacts upon entering the execute phase for your research_id and enforces their use to validate and control writes throughout the evidence-loop workflow.

The best way to manage loop progression in hydrology research workflows is by invoking the HydroMind command to enter the execute phase for your research_id. The system loads required artifacts and enforces sequential loop progression.