What problem does it solve?
After a main experiment produces a result, research quests often need structured follow-up runs such as ablations, robustness checks, sensitivity tests, or failure analysis. Without a disciplined protocol, these runs become an untracked pile of extra experiments with buried negative results and no clear connection back to the paper or claim they support.
Core Features & Use Cases
- Structured Campaign Protocol: Launches each campaign with a durable decision artifact, a charter report, a PLAN.md, and a CHECKLIST.md so every slice is planned, tracked, and recorded.
- Paper-Facing Slice Binding: Binds each analysis slice to a selected outline, a paper experiment matrix, and fields like section_id, claim_links, and paper_role so results map directly into manuscript sections.
- Comparability and Honest Reporting: Enforces one question per run, fixed evaluation contracts, visible null and negative findings, and an aggregated report classifying outcomes as stable support, partial support, contradiction, or unresolved ambiguity.
- Use Case: After a main experiment shows your method beats a baseline, use this Skill to run a claim-critical ablation, a seed-robustness check, and a failure-bucket analysis, then aggregate the evidence and route the quest to paper writing.
Quick Start
Run an analysis campaign on my main experiment with ablations for each core component and a robustness check across three random seeds, then summarize whether the main claim still holds.
