sci-review
Specialized workflows for drafting, refining, and responding to academic literature reviews and peer review feedback. Triggers on requests like "draft literature review on...", "respond to reviewers", "refine paper rebuttal", or "summarize research gaps". Based on the 4-stage systematic literature review structure and professional rebuttal guides for top-tier CS venues (NeurIPS, ICLR, ICML, etc.).
Sci-Review — Scientific Review & Rebuttal Skill
This skill consolidates professional academic writing logic to help authors create high-impact literature reviews and effective responses to peer reviewers.
Core Capabilities
1. Literature Review Orchestration
Follows a rigorous 4-stage structure optimized for high-tier journals (JCR Q1+ standards):
- Introduction: Background -> Problem Definition -> Gap Identification -> Contribution.
- Methodology/Materials: Systematic classification, cross-method comparison, and performance benchmarking.
- Challenges (3-Layer structure): Phenomenon (What is wrong) -> Cause (Why it is wrong) -> Direction (How to fix it).
- Conclusion: Distilled insights and future roadmap.
2. Professional Rebuttal Drafting
Transforms author responses from defensive arguments into professional technical dialogues:
- Tone Control: Replaces "reviewer misunderstood" with "we will clarify in the text".
- Evidence-Driven: Prioritizes numerical data, paper quotes, and specific section references.
- Structure: Systematic addressing of factual errors, missing experiments, and conceptual disagreements.
Usage Guide
Drafting a Review
When asked to start a review:
- Define Taxonomy: Establish 5-8 clear dimensions for classification.
- Comparative Analysis: Always use evidence-based comparisons (numerical data over vague descriptors).
- Future Outlook: Ensure the "Challenges" section follows the Phenomenon-Cause-Direction hierarchy.
Writing a Rebuttal
When asked to respond to reviewers:
- Quoting: Quote the core of the reviewer's concern accurately.
- Direct Action: Each response should ideally result in a concrete commitment to revise the manuscript.
- Word Economy: Focus on major concerns (Reviewer scores/Gatekeeper comments) first.
Best Practices
- Read First: Always read the source literature or reviewer comments thoroughly before generating content.
- No Fillers: Remove phrases like "it is worth noting" or "as we all know".
- Precision: Replace "significantly better" with "X% improvement over [Baseline]".
© License & Copyright
Aut_Sci_Write — Autonomous Scientific Writer
- Author: Shuo Zhao
- License: MIT License
- Copyright: © 2026 Shuo Zhao. All rights reserved.
- Original Work: This is an original work created by the author. No reproduction, redistribution, or commercial use without explicit permission. Permission is hereby granted, free of charge, to any person obtaining a copy of this software... (See the LICENSE file in the root directory for the full MIT terms.)