Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› rfs-rebuttal
rfs-rebuttal
GitHub用于RFS期刊修改重投（R&R）或拒稿重投场景，协助构建回复信及修订计划。指导如何结构化回应审稿意见、处理不同评论类型，并强调代码发布与注册报告规范，确保回复专业且严谨。
Trigger Scenarios
收到RFS修改重投决定信
需要制定修订和回复策略
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill rfs-rebuttal -g -y
SKILL.md
Frontmatter
{
"name": "rfs-rebuttal",
"description": "Use after a decision letter (R&R or reject-and-resubmit) on a The Review of Financial Studies (RFS) manuscript, to structure the response letter and revision plan. Writes the rebuttal; assumes the manuscript itself is being revised in parallel."
}
R&R Rebuttal & Response Letter (rfs-rebuttal)
When to trigger
- An RFS decision letter arrives with an editor letter and referee reports
- You have a revise-and-resubmit (or a "reject and resubmit") and must respond
- You need to plan which comments to fully address, partially address, or push back on
Revise the manuscript first (or in parallel). Do not draft the response letter around changes you have not actually made.
RFS revision reality
RFS runs a demanding multi-round review under a handling Editor (within the editor pool led by Executive Editor Tarun Ramadorai). An R&R is an opportunity, not an acceptance: the editor and referees expect substantive responses, new analysis where warranted, and a manuscript that is visibly better. The response letter is judged as carefully as the paper.
Two RFS-specific revision situations:
- Registered Report Stage 2. If the paper received Stage 1 in-principle acceptance, the Stage 2 review checks adherence to the pre-registered protocol, not whether results are "interesting." Your response letter must map each pre-specified analysis to its result and flag — transparently — any deviation and why. Adding unplanned analyses is fine but must be labeled exploratory, not slipped in as confirmatory.
- Code-release at revision. RFS makes public code release a condition of publication. Use the revision rounds to get the replication package clean; referees may re-run it. Do not defer this to post-acceptance.
Response-letter structure
- Cover note to the editor (half a page): thank the editor and referees; summarize the 3–4 most important changes; note any comment you respectfully disagree with and why. Address the editor's own priorities first — the editor decides.
- Per-referee, per-comment: reproduce each comment verbatim, then respond directly beneath it.
- For each comment use a consistent format:
- Comment (quoted)
- Response (what you did)
- Change (where in the revised paper/IA, with page/section/table numbers)
How to handle each comment type
- Substantive & correct → do the analysis; show the new result; point to the exact location.
- Substantive but you disagree → push back with evidence, politely; offer an additional test that addresses the underlying concern even if you keep your specification.
- Based on a misreading → clarify the text so the next reader cannot misread it too; thank the referee for revealing the ambiguity.
- Out of scope / would change the paper → explain the scope; offer to note it as a limitation or future work.
- Conflicting referees → surface the conflict to the editor and explain your chosen resolution.
Tone and discipline
- Respectful, specific, evidence-based. Never dismissive.
- Quote your own added text/results so the referee need not hunt for them.
- Use a colored/italicized revision-text convention so changes are easy to verify.
- Track every comment to closure — an unaddressed comment is the fastest path to rejection.
Checklist
- Editor cover note summarizes the major changes and addresses editor priorities first
- Every referee comment reproduced verbatim and answered individually
- Each response names the exact location of the change (page/section/table/IA)
- New analyses actually run and incorporated, not promised
- Disagreements are evidence-based and offer an alternative test
- Conflicting referee demands surfaced and resolved transparently
- Revision-text convention used so changes are easy to locate
- For a Registered Report: every pre-specified analysis mapped to result; deviations flagged; new analyses labeled exploratory
- Replication package updated and runnable (RFS code-release condition)
- No comment left unaddressed
Anti-patterns
- Promising changes in the letter that are not actually in the manuscript.
- Dismissing a referee or arguing without evidence.
- A vague "we have revised accordingly" with no location.
- Silently ignoring a comment you find inconvenient.
- Picking a fight with the editor's stated priorities.
- Over-adding analyses that bloat the paper instead of using the Internet Appendix.
Output format
【Decision】R&R / reject-and-resubmit / Registered Report Stage 2
【Editor priorities】[the 2–3 things the editor most wants]
【Comment ledger】[referee#-comment# → addressed / partial / pushed-back → location]
【New analyses】[run + incorporated; label exploratory vs. confirmatory if a Registered Report]
【Code package】updated + runnable? yes/no
【Open risks】[comments that may not satisfy]
【Status】response letter ready / manuscript revised? yes/no
Version History
- 1839142 Current 2026-07-05 14:23
