research-refine-pipeline

To build an end-to-end research pipeline, you provide a rough research direction and the system chains research-refine and experiment-plan workflows to output a final proposal and a ready-to-run experiment roadmap.

Chaining research refinement and experiment planning workflows generates gated artifacts including refine-logs, FINAL_PROPOSAL.md, EXPERIMENT_PLAN.md, and a PIPELINE_SUMMARY.md to document the entire pipeline.

Yes, you can connect multiple research workflows into a single one-shot pipeline that moves from a broad research direction to a detailed experiment roadmap while preserving the Problem Anchor throughout.

Yes, the research pipeline includes phase-driven governance steps such as triage, refinement, gating, and integration to ensure a robust method and final proposal before experimentation begins.

You refine a broad research direction by inputting it into the pipeline, which automatically performs method refinement and experiment planning to produce a concrete proposal and detailed experiment roadmap.

The one-shot workflow for research refinement and experiment planning requires a rough research direction as input and focuses on producing final artifacts, meaning it may not suit iterative, exploratory research processes needing manual intervention between phases.