paper-plan converts review conclusions, narrative reports, and experiment outputs into a detailed, section-by-section paper outline. It invokes a local Gem…
paper-plan converts review conclusions, narrative reports, and experiment outputs into a detailed, section-by-section paper outline. It invokes a local Gemini reviewer (REVIEWER_MODEL = gemini-review, with optional GEMINI_REVIEW_MODEL override), honors TARGET_VENUE (ICLR/NeurIPS/ICML) and MAX_PAGES constraints, and automatically extracts 3–5 core claims, linked evidence (experiments, metrics, figures), known weaknesses, and suggested framing. The skill builds a Claims–Evidence Matrix mapping each claim to supporting data and target sections, classifies paper type, and generates a page-aware structure (Intro, Related Work, Methods, Experiments, Limitations, Conclusion, etc.). Use when you have NARRATIVE_REPORT.md, GPT54_AUTO_REVIEW.md, experiment JSONs/figures, or can supply a 3–5 sentence contribution summary. Core advantages: reviewer-informed framing, explicit evidence-to-section mapping, and venue/page-limit aware outlines to accelerate paper drafting.
This page belongs to the OpenClaw Skills learning hub with install guides, category navigation, and practical links.