paper-referee-revise
Revise an academic paper based on journal referee reports. Reads referee comments from review report or annotated manuscript, then directly modifies main.tex one comment at a time with user approval. Generates response letter after revision.
Install / Use
npx skills add brycewang-stanford/Auto-Empirical-Research-Skills --skill paper-referee-reviseInstalls into whichever agent you are using.
SKILL.md
Installable skill definition
Quality Score
Category
MarketingSupported Platforms
Our assessment of paper-referee-revise
paper-referee-revise scores 95/100 on our quality scale, 79th of 642 Marketing skills we index (top 13%).
Its SKILL.md is 26 KB long, well organised into 23 sections with 11 code examples: a thorough specification that gives an agent plenty to work with.
With 4,360 GitHub stars, it is one of the more widely adopted skills in the catalogue.
Maintenance, license and trust
- The repository was last updated about 2 months ago, so paper-referee-revise is actively maintained.
- Our last check on 2026-11-15 found the source still online.
- No license is declared. By default that means all rights are reserved: you can read it, but reusing or redistributing it is not clearly permitted. Ask the author before building on it commercially.
- Its trust signals score 88/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-09-29. Automated pattern scan on 2026-09-29. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
paper-referee-revise compared with similar skills
All 4 of these similar skills score higher than paper-referee-revise; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| paper-referee-revise (this skill)by brycewang-stanford | 95 | 4.4k | 54d ago | SKILL.md |
| x402by browser-use | 100 | 121.0k | 3d ago | SKILL.md |
| algorithmic-artby anthropics | 100 | 177.9k | 55d ago | SKILL.md |
| pptxby anthropics | 100 | 177.9k | 55d ago | SKILL.md |
| designby nextlevelbuilder | 100 | 149.0k | 6d ago | SKILL.md |
Frequently asked questions
- How do I install paper-referee-revise?
- Run
npx skills add brycewang-stanford/Auto-Empirical-Research-Skills --skill paper-referee-revise. The install tabs above show the steps for each supported agent. - Which AI agents does paper-referee-revise work with?
- It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
- Is paper-referee-revise safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It declares no license and scores 88/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is paper-referee-revise still maintained?
- The repository was last updated about 2 months ago, so paper-referee-revise is actively maintained.
Skill content
View source on GitHubname: paper-referee-revise description: Revise an academic paper based on journal referee reports. Reads referee comments from review report or annotated manuscript, then directly modifies main.tex one comment at a time with user approval. Generates response letter after revision. Trigger when user says "referee revise" / "paper-referee-revise" / "审稿意见修改" / "根据审稿人意见修改" / "referee report". allowed-tools: Read, Edit, Write, Glob, Grep, AskUserQuestion, Bash argument-hint: "[path-to-project-folder]"
Paper Referee Revise
Overview
This skill revises an academic paper (main.tex) based on journal referee reports received during the peer review process. The user provides a project folder containing the manuscript, supporting files, and referee materials. The skill extracts referee comments, triages empirical work, then processes text revisions one by one with user approval, and finally drafts a response letter.
Input: A project folder containing:
main.tex— the main manuscript (the source file; will NOT be overwritten)- Subfolders with figures/tables (referenced by the manuscript)
- Referee materials (one or more of the following):
- A referee report file (e.g.,
referee.txt,referee.md,referee.pdf,report.txt,report.md,report.pdf,review.txt,review.md,review.pdf,R1.txt,R1.pdf,referee1.*,referee2.*) - An annotated manuscript (e.g.,
annotated.tex,annotated.pdf) where annotations/comments from the referee serve as review comments - Multiple referee reports (e.g.,
referee1.pdf,referee2.pdf) — process each referee's comments separately
- A referee report file (e.g.,
Output:
- If empirical comments exist: generates
empirical-revise.pdf(empirical revision strategy) and halts. - If no empirical comments (or empirical work already done): generates
main-rX.tex(revised manuscript with changes in red) andresponse-rX.tex(response letter), then halts.
Workflow
Phase 1: Initialization
- Receive folder path from user (via
$ARGUMENTSor ask). - Ask the user: "本次修改为第几轮修改?(Please enter the revision round number, e.g., 1, 2, 3)" Store the answer as
X— this determines output file names:main-rX.texandresponse-rX.tex. - Determine the source manuscript based on revision round:
- If
X = 1: the source file ismain.tex. - If
X = 2: the source file ismain-r1.tex. - If
X = 3: the source file ismain-r2.tex. - General rule: for round X, the source file is
main-r{X-1}.tex(i.e., the output of the previous round). IfX = 1, usemain.tex. - Verify the source file exists. If it does not (e.g., user says round 2 but
main-r1.texis missing), alert the user and ask them to confirm.
- If
- Scan the folder using Glob to identify all files:
- Locate the source manuscript file (as determined in step 3 — read only, never overwrite).
- Locate referee materials by searching for files matching patterns:
referee*,report*,review*,R1*,R2*,R3*,annotated*,comments*,feedback*,revision*. - If multiple referee report files are found, list them and ask the user which one(s) to use and in what order.
- If no referee file is found, ask the user to specify the file path.
- Read the source manuscript in full to understand the current state of the paper.
- Read all table and figure
.texfiles from subfolders so that you have full knowledge of the paper's empirical results. This is critical for proposing accurate revisions when referees comment on results. - Read the referee file(s):
- For
.texfiles: look for comments marked with%or annotation commands (e.g.,\annotate{},\comment{},\todo{}, or any custom annotation commands, or highlighted/colored text added by the referee). - For
.txt/.mdfiles: read the full content as referee comments. - For
.pdffiles: read the PDF content to extract referee text. - For
.docxfiles: use Bash to extract text content.
- For
Phase 2: Extract and Structure Referee Comments
- Identify how many referees provided reports (Referee 1, Referee 2, etc.). If the report contains comments from multiple referees, separate them.
- Parse the referee materials to extract individual comments/suggestions from each referee.
- Classify each comment by type:
- Empirical: Requires new regressions, robustness checks, new tables/figures, or modifications to existing empirical results
- Text: Requires only text changes to the manuscript (clarification, rewording, restructuring, adding discussion)
- Editorial: Typos, grammar, formatting issues
- Structure each comment as a discrete revision item with:
- Comment ID (e.g., R1-1, R1-2, R2-1 for Referee 1 Comment 1, etc.)
- Type (Empirical / Text / Editorial)
- Referee's original comment text
- Location in manuscript (section, paragraph, or specific text the comment refers to)
- Present the full list of comments to the user, grouped by referee:
Referee 1 comments:
ID | Type | Location | Comment Summary
------|-----------|---------------------------|------------------------------------------
R1-1 | Empirical | Section 3, para 2 | Endogeneity concern, suggests IV approach
R1-2 | Text | Section 4.1, Table 2 | Clarify discussion of control variables
R1-3 | Editorial | Section 2, para 4 | Typo in variable name
Referee 2 comments:
ID | Type | Location | Comment Summary
------|-----------|---------------------------|------------------------------------------
R2-1 | Empirical | Section 5 | Additional robustness checks needed
R2-2 | Text | Introduction, para 3 | Clarify contribution statement
Total: N comments (X empirical, Y text, Z editorial).
- Wait for user confirmation before proceeding.
Phase 3: Empirical Triage
This phase is mandatory. It runs before any text revision begins.
-
Identify empirical comments: From Phase 2 classification, collect all comments typed as "Empirical" — i.e., any comment that demands:
- New regressions or estimations (e.g., new dependent variables, new specifications, IV/2SLS/GMM, DID, RDD, etc.)
- Additional robustness checks (e.g., alternative measures, subsample analysis, placebo tests, different fixed effects, winsorizing, clustering)
- New summary statistics or descriptive analysis
- New tables or figures not currently in the project folder
- Re-running existing regressions with modifications (e.g., adding/removing control variables, changing sample period)
- Additional tests (e.g., Hausman test, overidentification, weak instrument tests, parallel trends, balance tests)
- Data-related requests (e.g., new data sources, extended sample, different frequency)
-
If empirical comments exist, generate
empirical-revise.texin the project folder with the following structure:
\documentclass[12pt,a4paper]{article}
\usepackage[utf8]{inputenc}
\usepackage[margin=1in]{geometry}
\usepackage{enumitem}
\usepackage{booktabs}
\usepackage{longtable}
\usepackage{xcolor}
\title{Empirical Revision Strategy}
\author{}
\date{}
\begin{document}
\maketitle
\section{Overview}
This document summarizes all empirical work required to address referee comments.
Total empirical comments: N (out of M total comments).
\section{Empirical Tasks}
\subsection{[R1-1] [Brief title of the empirical task]}
\textbf{Referee's comment:} [Full text of the referee's original comment]
\textbf{Current status:} [What the paper currently does — reference existing table/figure if applicable]
\textbf{Required action:}
\begin{itemize}
\item [Specific step 1: e.g., "Run OLS regression of Y on X with firm and year FE, adding control variables Z1, Z2"]
\item [Specific step 2: e.g., "Store results in a new table: table_iv.tex"]
\item [...]
\end{itemize}
\textbf{Suggested table/figure name:} [e.g., table\_iv.tex, table\_robust\_subsample.tex]
\textbf{Suggested placement:} [e.g., "Section 4.2, after Table 3" or "Appendix"]
% Repeat \subsection for each empirical comment
...
\section{Summary of New Tables/Figures Needed}
\begin{longtable}{lll}
\toprule
File name & Description & Placement \\
\midrule
table\_iv.tex & IV/2SLS results for endogeneity & Section 4.2 \\
table\_robust\_sub.tex & Subsample robustness & Section 5.1 \\
... & ... & ... \\
\bottomrule
\end{longtable}
\section{Non-Empirical Comments (For Reference)}
The following comments do not require new empirical work and will be addressed in the text revision phase:
\begin{itemize}
\item [R1-3] [Brief summary]
\item [R2-2] [Brief summary]
\item ...
\end{itemize}
\end{document}
- Compile to PDF: Run via Bash:
cd [project-folder] && pdflatex empirical-revise.tex
If pdflatex fails, try xelatex empirical-revise.tex. If compilation fails entirely, inform the user that the .tex file is ready and they can compile it manually.
- HALT the program. Display the following message and stop:
══════════════════════════════════════════════════════════════
实证修改策略已生成: [project-folder]/empirical-revise.pdf
共发现 N 条涉及实证的审稿意见,需要先完成以下工作:
- 新增表格: X 个
- 修改现有表格: Y 个
- 新增图表: Z 个
请完成实证工作后,将新的表格/图表文件放入项目文件夹,
然后重新运行本skill继续修改论文正文。
══════════════════════════════════════════════════════════════
Do NOT proceed to Phase 4. The program ends here.
- If NO empirical comments exist, skip this phase entirely and proceed directly to Phase 4.
Re-entry after empirical work is done: When the user re-runs this skill on the same folder after completing the empirical work, Phase 3 should detect that empirical-revise.pdf already exists. Ask the user to confirm: "实证修改是否已全部完成?(Has all empirical work been completed?)" If the user confirms yes, proceed to Phase 4. If not, the user should complete the remaining work first.
Phase 4: Iterative Revision (Core Loop)
Before starting: Copy the source manuscript to main-rX.tex (i.e., for round 1 copy main.tex → main-r1.tex; for round 2 copy main-r1.tex → main-r2.tex; etc.). All subsequent edits are made to main-rX.tex, NEVER to the source file. Ensure main-rX.tex includes \usepackage{xcolor} in the preamble (add it if not already present) — this is required for red markup. If the source file already contains \textcolor{red}{} markup from a previous round, convert those red markings to normal black text first (remove the \textcolor{red}{} wrappers, keeping the content), so that only the current round's changes appear in red.
For each referee comment (both text-only AND empirical comments that now have completed results), follow this exact interaction pattern:
Step 1: Display the referee comment
───────────────────────────────────────────
[R1-1] (Empirical/Text/Editorial) — Referee 1, Comment 1:
[Full text of the referee's comment]
───────────────────────────────────────────
Step 2: Show original text
Locate and display the relevant passage from main-rX.tex:
Original text (main-rX.tex, lines XX-YY):
┌─────────────────────────────────────────┐
│ [The exact paragraph(s) from main-rX.tex│
│ that this comment refers to] │
└─────────────────────────────────────────┘
If the comment refers to missing content (e.g., "the authors should add a discussion of X"), show the surrounding context where the new text should be inserted.
Step 3: Propose revision
Generate a revised version of the text that addresses the referee's comment. All new or changed text must be wrapped in \textcolor{red}{...} so that changes are visually highlighted in the compiled PDF:
Proposed revision:
┌─────────────────────────────────────────┐
│ [The revised paragraph with changed │
│ portions wrapped in \textcolor{red}{}] │
└─────────────────────────────────────────┘
Rationale: [Brief explanation of what was cha
Truncated for display — read the full file on GitHub.
Related Skills
x402
121.0kSet up Browser Use Cloud payments with x402 — pay per request from a crypto wallet (USDC on Base mainnet), no signup or API key.
algorithmic-art
177.9kCreating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems.
pptx
177.9kUse this skill any time a .pptx or .potx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx or .potx file (even if the extracted content will be used elsewhere, like in an em…
design
149.0kComprehensive design skill: brand identity, design tokens, UI styling, logo generation (55 styles, Gemini, Atlas Cloud, or MuAPI AI), corporate identity program (50 deliverables, CIP mockups), HTML presentations (Chart.js), banner design (22 styles, social/ads/web/print), icon design (15 styles, SVG…
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
