Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

修改回复（cre-rebuttal） · brycewang-stanford | DecimalAI

brycewang-stanford/cre-rebuttal

SkillScore55Maintained

Use when responding to reviewer reports for a 《中国农村经济》 R&R. Produces the point-by-point response letter and aligned manuscript edits. 本技能服务于《中国农村经济》(China Rural Economy, CRE)。

Documentationby brycewang-stanford

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Research pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/brycewang-stanford-cre-rebuttal/SKILL.md

decimalai skills pull brycewang-stanford-cre-rebuttal --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/brycewang-stanford-cre-rebuttal/SKILL.md https://app.decimal.ai/s/brycewang-stanford-cre-rebuttal/SKILL.md

Raw file: /s/brycewang-stanford-cre-rebuttal/SKILL.md ↗ · machine-readable summary at /s/brycewang-stanford-cre-rebuttal.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

Runs

all versions · 30d

Eval pass

100%

global

Used when offered

100%

used / offered · all versions

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 5d ago

Yes — +36% lift · at higher cost+58% tokens0% turns55% of whole cases pass with skill (12/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-13	✗→✓	▲ Improved	145%	0%
case-04	✗→✓	▲ Improved	61%	0%
case-06	✗→✓	▲ Improved	65%	0%
case-07	✗→✓	▲ Improved	73%	0%
case-11	✗→✓	▲ Improved	86%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

brycewang-stanford / awesome-journal-skills

China-Rural-Economy-Skills/skills/cre-rebuttal/SKILL.md

View on GitHub ↗

License MITImported 8/10/2026Version v1

Skill contents

SKILL.md

修改回复（cre-rebuttal）

触发时机

收到《中国农村经济》R&R 通知
已有 1–3 份匿名审稿意见和编辑部 / 编委意见
准备开始修订正文 + 回复信

顺序原则

先修订正文，再写回复信。回复信是"导览图"，必须基于已修订的正文版本。

颠倒会导致：

回复信承诺了正文未真正做的修订
正文修订过程中发现承诺无法兑现
多稿来回不一致

回复信结构

总体结构

尊敬的主编、编委、各位审稿专家：

非常感谢编辑部和各位审稿专家在百忙之中对本文的审阅和宝贵意见。
我们已根据各位意见对论文进行了系统修订。下面我们逐条回复各位的意见。
为便于核对，我们在回复中标注了修订对应的正文位置（如"详见正文 P.X 第 Y 段"）。

致主编 / 编委的回复（如有编委意见）
审稿人 1 的回复
审稿人 2 的回复
审稿人 3 的回复（如有）

再次感谢各位的宝贵意见。
作者
20XX 年 XX 月 XX 日

单条回复结构

意见 X：[原文照抄审稿意见]

回复：感谢审稿专家的宝贵意见。我们已[做了什么具体修改]，理由如下：
（1）从理论角度，……
（2）从实证角度，……
（3）从三农场景角度，……
具体修改详见正文第 X 页第 Y 段（修订标记已用[蓝色字体 / 修订模式]标出）。

三种回复模式

A. 完全接受

直接说明接受 + 描述修改 + 指明位置
不要找理由

B. 部分接受

先肯定审稿人的合理性
说明部分接受的范围
解释未接受部分的原因（给证据，不要狡辩）

C. 不接受（谨慎使用）

先充分肯定审稿人的考虑
给出严密的反驳论证（至少 3 段）
必要时新增稳健性 / 补充分析作为旁证
不接受的比例最多 1–2 条

《中国农村经济》R&R 特别注意

审稿人最常要求的三类补充：(1) 农户自选择内生性的进一步处理（补 IV / 安慰剂 / 敏感性）；(2) 机制落到农户层面（不要停在宏观叙事）；(3) 异质性维度贴合三农场景
若审稿人质疑"识别不干净"，应额外加 2–3 个稳健性表格附在文末（更换样本、更换识别方法、安慰剂）
若审稿人要求"补充机制" → 优先在正文加机制小节，而非附录
若编辑部 / 编委意见关于三农场景契合度或政策含义，必须重点回应——这是本刊定位的核心
数据相关质疑（调查代表性、变量口径）要正面交代抽样与构造，不可回避

必查清单

] 每条审稿意见都有对应回复（不能遗漏）
] 每条回复都指明正文修订位置
] 回复信开头有致谢段
] 不接受的意见有充分论证 + 旁证
] 正文修订处有标记（蓝色字体 / 修订模式）
] 修订版正文已通读，确保无前后矛盾
] 三农场景契合度 / 政策含义类意见已优先回应
] 自选择内生性、机制、异质性类意见已落实到正文表格

反模式

"感谢审稿人的意见，我们已修改"——一行回复
回复信比正文还长 —— 多半在凑字数
三个审稿人意见冲突，回复信里全部接受 —— 正文会自相矛盾
把不接受的回复写得很 defensive
忽视编辑部 / 编委意见，只回复匿名审稿人

输出格式

【审稿人数】X 人
【意见总数】X 条
【完全接受】X 条
【部分接受】X 条
【不接受】X 条
【新增稳健性 / 机制表格】X 张
【正文修订段落数】X
【编委意见回应】完整 / 待补
【下一步】上传至投稿系统 → 等待二轮审稿

《中国农村经济》操作审查

先锁定农村问题、政策/制度场景、识别链条、机制证据和可执行含义，再判断稿件是否回应农村经济审稿人通常同时追问“三农”问题意识、政策场景、识别可信度和农村制度机制。

回应审查：按编辑优先级和审稿人编号分开处理，每条重大意见必须对应正文修改、表图证据或附录补充。
决策账本：返回“主张 / 证据 / 阻断点 / 下一处改稿”四列，便于下一轮直接改稿。
改投比较：对照《经济研究》用于更一般理论贡献，《管理世界》用于治理/管理实践，《农业经济问题》用于更专门农业政策；若相邻刊物拥有更强读者匹配，先建议改投而不是继续润色。
核验底线：给投稿就绪判断前，必须重开 resources/official-source-map.md，列出仍可能改变建议的一个未核实事实。

Want to see if brycewang-stanford/cre-rebuttal would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/31/2026, 05:48 UTC · 22 cases

Lift

+36%

55% vs 18%

Agent turns

1.0 vs 1.0

Tokens

+58%

4,496.27 vs 2,837.86

fail → pass · 8pass → pass · 4fail → fail · 10

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 我刚收到了《中国农村经济》期刊退修重审（R&R）的通知，包含了编委意见和两位审稿人的修改建议，主要集中在农户自选择内生性处理和三农政策含义贴合度上。请帮我梳理修改策略，并撰写一份符合该刊规范的逐条回复信草稿。要求包含开头致谢、各审稿人意见的分类回应以及正文修改定位提示，并在末尾提供包含审稿人数、意见分类计数与后续步骤的统计数据块。	fail→fail	36,998	36,102	-2%	1	1	0%	4,868	6,297	+29%	0	0	—
case-02 我们有一篇关于农村集体经济的论文收到了《中国农村经济》的退修意见，其中有一条关于样本筛选的意见我们认为不合理、无法接受。请帮我拟定一份完整的回复方案：针对不合意见给出严密且有证据支撑的反驳，整理出一份包含主张、证据、瓶颈与改稿动作的四列决策表格，并附上最终修改情况的汇总统计。	fail→fail	29,147	28,100	-4%	1	1	0%	3,670	4,723	+29%	0	0	—
case-03 投稿《中国农村经济》的稿件拿到了三位审稿人的意见，要求补充农户微观机制和稳健性检验。请按照该期刊的标准要求，为我规划正文修改与回复信撰写的先后流程，并生成一份包含编委及各审稿人逐条回应的完整回复文档，同时在文末列出修订意见分布及提交状态摘要。	fail→fail	37,973	46,393	+22%	1	1	0%	5,340	7,249	+36%	0	0	—
case-08 Reviewer 2 requested that we re-estimate our empirical model across all 31 provinces, but our survey dataset covers only 10 major grain-producing provinces. We want to accept the comment for grain provinces but cannot expand to non-grain provinces. Which response mode applies?	fail→fail	18,541	23,620	+27%	1	1	0%	1,862	4,158	+123%	0	0	—
case-13 We finished addressing all reviewer comments for our Chinese Rural Economy paper. Please produce the standardized revision metadata block detailing reviewer counts, decision classifications, and next action.	fail→pass	9,027	16,944	+88%	1	1	0%	1,281	3,136	+145%	0	0	—
case-04 I am preparing the R&R package for an agricultural policy paper. My co-author suggests drafting the response letter first so we know what to change in the paper later. Should we write the response letter first or modify the main text first? Please explain the workflow rule and outline the single-item response template.	fail→pass	16,132	26,015	+61%	1	1	0%	2,439	3,935	+61%	0	0	—
case-05 A reviewer asked us to drop our difference-in-differences identification strategy and switch to simple OLS. We strongly disagree because OLS fails to control unobserved village-level heterogeneity. How should we write this disagreement, and how many comments can we reject?	pass→pass	20,361	29,769	+46%	1	1	0%	2,425	4,919	+103%	0	0	—
case-06 The reviewer commented that our farm household adoption model suffers from self-selection endogeneity. My co-author wants to write a text-only paragraph arguing that self-selection is minimal. Should we accept this text-only approach or make econometric additions to the paper?	fail→pass	22,916	27,561	+20%	1	1	0%	2,616	4,324	+65%	0	0	—
case-07 Our paper analyzes rural land transfer policy using county-level aggregate data. Reviewer 1 requests that we demonstrate the micro-level mechanism. Can we satisfy this by adding a conceptual narrative to the introduction section?	fail→pass	21,825	27,191	+25%	1	1	0%	2,591	4,495	+73%	0	0	—
case-19 We are preparing an initial first-round submission to Chinese Rural Economy. What are the journal's required font size, margin width, and reference formatting rules for new submissions?	fail→fail	26,503	20,765	-22%	1	1	0%	3,836	4,802	+25%	0	0	—
case-09 Our paper on agricultural supply chain governance got an R&R, but reviewer feedback suggests our main contribution is general transaction cost economics rather than specific rural policy mechanisms. Should we stay with Chinese Rural Economy or consider redirecting (改投) to other top journals?	fail→fail	22,975	31,269	+36%	1	1	0%	2,700	5,347	+98%	0	0	—
case-10 We received R&R comments from two anonymous reviewers and one editorial board member. The reviewers requested minor control variable tweaks, while the board member requested strengthening the paper's alignment with national 'three-rural' (三农) policy context. Which feedback takes priority?	fail→fail	18,650	25,585	+37%	1	1	0%	2,170	4,477	+106%	0	0	—
case-11 Reviewer 2 questioned whether our 500-household survey in Henan province has sufficient sampling representativeness and criticized our variable definition of off-farm labor. Should we avoid answering this data query directly?	fail→pass	19,215	20,908	+9%	1	1	0%	2,352	4,385	+86%	0	0	—
case-12 We need to construct a structured decision ledger table for our co-authors to track 8 revision points for a Chinese Rural Economy R&R. What columns should this decision table contain?	fail→pass	22,466	19,054	-15%	1	1	0%	2,478	4,084	+65%	0	0	—
case-14 Reviewer 1 asked us to remove variable X to avoid over-controlling, while Reviewer 2 insisted that variable X must remain to prevent omitted variable bias. Should we accept both suggestions in full?	pass→pass	19,292	26,074	+35%	1	1	0%	2,065	4,021	+95%	0	0	—
case-15 We are ready to submit our second-round R&R response package to Chinese Rural Economy. What final source-verification baseline check must be performed before declaring the paper ready?	fail→pass	23,290	15,624	-33%	1	1	0%	2,401	3,017	+26%	0	0	—
case-16 What mandatory items must we verify on our final checklist before uploading our revised manuscript and rebuttal letter to Chinese Rural Economy?	fail→pass	19,033	8,758	-54%	1	1	0%	2,616	2,835	+8%	0	0	—
case-17 My co-author drafted a 50-page rebuttal letter for a 20-page article in Chinese Rural Economy, arguing defensively against Reviewer 1's comments. What formatting and tone errors should we fix?	pass→pass	27,963	26,424	-6%	1	1	0%	2,889	4,153	+44%	0	0	—
case-18 Reviewer 3 requested heterogeneity analysis for our digital agriculture paper in Chinese Rural Economy. My student proposed splitting the sample by respondent gender and age. Is this appropriate for this journal?	fail→fail	17,888	25,982	+45%	1	1	0%	2,597	4,041	+56%	0	0	—
case-20 I was invited by Chinese Rural Economy to serve as a peer reviewer for a submitted manuscript on rural land rights. Please write a formal peer review report evaluating the paper's contribution and methodology.	fail→fail	21,430	24,552	+15%	1	1	0%	2,826	4,739	+68%	0	0	—
case-21 We received a major revision decision from Nature Food on our global agricultural supply chain paper. Please draft an English point-by-point response letter following Nature portfolio response formats.	fail→fail	33,541	29,907	-11%	1	1	0%	3,884	5,919	+52%	0	0	—
case-22 I am working on an econometrics assignment for my graduate economics course. How do I calculate the Hausman specification test statistic to compare fixed effects and random effects models in Stata?	pass→pass	14,981	13,389	-11%	1	1	0%	2,527	3,862	+53%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +36 percentage points is the difference between those two pass rates over the 22 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 18.2%
Pass rate with the skill: 54.6%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 0
Skill version measured: v1
Run completed: 8/31/2026, 05:48 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-13

Prompt

We finished addressing all reviewer comments for our Chinese Rural Economy paper. Please produce the standardized revision metadata block detailing reviewer counts, decision classifications, and next action.

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run55

Live evalNeeds ≥20 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+71% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 5 days ago

Other measured skills in the registry, with their headline benchmark lift.