What problem does it solve?
This Skill converts your review conclusions and experiment results into a coherent, section-by-section paper outline so you can move from findings to a submission-ready structure faster.
Core Features & Use Cases
- Claims-to-outline planning: Builds an evidence-backed outline by extracting claims, mapping them to experiments, and positioning them within sections.
- Venue-aware page budgeting: Adjusts structure and length targets for common ML venues like ICLR, NeurIPS, ICML, and IEEE formats.
- Figure and citation scaffolding: Plans hero Figure 1 plus tables/figures, and provides a citation plan by section to reduce drafting friction.
- Optional style reference control: When you pass a style reference, it uses it for structural guidance (not prose copying) to steer layout density and planning.
Quick Start
Ask the AI to generate a paper outline for your work by saying: "写大纲 for my paper using my NARRATIVE_REPORT.md and AUTO_REVIEW.md, target venue ICLR."