Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› aos-rebuttal
aos-rebuttal
GitHub用于处理AOS期刊修改重投（R&R）的回复信撰写与修订规划。指导建立响应矩阵，区分重写、重分析等策略；针对定性或实验研究提供具体修订建议；协调跨学科审稿人冲突，确保逐点回应且语气专业，强调实质性数据/理论工作优先于文字修饰。
Trigger Scenarios
收到AOS期刊的修改重投（R&R）意见
需要制定包含重新编码或新实验条件的修订计划
审稿人意见存在跨学科冲突需协调
准备逐点回应审稿人和编辑的意见
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill aos-rebuttal -g -y
SKILL.md
Frontmatter
{
"name": "aos-rebuttal",
"description": "Use when an Accounting, Organizations and Society (AOS) revise-and-resubmit arrives — planning revisions that may include recoding field data, new experimental conditions, or deepened theorizing, and drafting the point-by-point response to an interdisciplinary reviewer panel. Drafts the response; it does not run new analysis (aos-data-analysis) or the final preflight (aos-submission)."
}
R&R Rebuttal (aos-rebuttal)
When to trigger
- An AOS R&R letter is in hand and the revision needs a plan
- Reviewers demand deeper theorizing, more field evidence, or a new experimental condition
- Two referees pull in different disciplinary directions and the response must reconcile them
- The response letter is due and the manuscript changes must be documented point by point
Plan the revision before a word of the letter
- Build a response matrix: every numbered point from each reviewer and the editor, mapped to one of — rewrite (clarity), re-analyze (return to the data), extend (new material), or reasoned dissent (rare, sourced).
- Let the editor's letter set the order of work. At AOS the editor typically synthesizes which concerns are decisive; treat that synthesis as the revision's spine.
- Do the substantive work first. AOS revisions frequently mean real labor, not prose: another coding pass with a sharpened category system, a follow-up interview round if access allows, a new experimental condition or process measure, or rebuilding the theory section from primary sources.
Tradition-specific revision moves
- Qualitative field papers. "The theory does no work" → re-theorize from the anomalies, not by adding citations; show in the response which findings paragraphs changed. "Evidence is thin" → expand the data-inventory and evidence tables, add counter-evidence handling; only claim saturation if the register supports it.
- Experiments. "Confound X" → new condition or measured-covariate analysis, not verbal reassurance. "Process unclear" → add mediation/moderation evidence with bootstrap CIs, or run the discriminating study.
- Cross-discipline conflicts. When the theory reader and the accounting reader collide, write the bridging move into the manuscript (a paragraph translating the theoretical stake into the accounting conversation), then point both reviewers to it — never give incompatible private answers in the letter.
Write the response letter
- Open with a one-page summary of the revision's architecture: the two or three big changes and where they live.
- Then point by point: quote or faithfully paraphrase the comment, state what changed, and give the exact location (section, page, table). Never claim a change the manuscript does not contain.
- Dissent sparingly and generously: acknowledge the reading, explain with sources or evidence why the paper holds its position, and offer the compromise made (a scope condition, a footnote, a robustness exhibit).
- Keep tone collegial — the same referees will likely see round two, and interdisciplinary reviewers are alert to being condescended to about "their" theory.
Consistency sweep before resubmission
- Re-run the terminology ledger: revisions notoriously desynchronize construct names between new sections and old exhibits.
- Update the data-inventory/evidence tables (qualitative) or the analysis scripts and exhibits (quantitative) so every changed number regenerates.
- Refresh declarations if anything moved: new data collection → ethics statement; AI-assisted redrafting → AI declaration; new related publication of yours → disclosure.
- Confirm the revised files remain fully anonymized — new acknowledgments and tracked-changes metadata are classic leaks.
Timing
Interpretive revisions are slow by nature — recoding a corpus or returning to a field site takes months, not weeks. If the editor's deadline cannot accommodate the substantive work the letter demands, ask for an extension early with a concrete plan; AOS editors generally prefer a real revision late to a cosmetic one on time (confirm the journal's current extension practice with the handling editor).
Checklist
- Response matrix covers every point with an action class
- Editor's decisive concerns addressed first and most deeply
- Substantive work (recoding / new conditions / re-theorizing) done before drafting
- Cross-reviewer conflicts resolved in the manuscript, identically for both readers
- Every response cites the exact changed location; no phantom changes
- Terminology, exhibits, declarations, and anonymization re-verified after revision
Anti-patterns
- Letter-first revising — eloquent responses wrapping an unchanged manuscript.
- Citation appeasement: adding the reviewer's references without engaging them.
- Splitting answers: telling R1 one thing and R2 its opposite.
- Silent dropping of a hard point, hoping the reviewer forgets.
- Over-compliance that dismembers the paper's argument to satisfy every marginal note.
Output format
【Matrix】R1 / R2 / editor points → rewrite / re-analyze / extend / dissent ...
【Spine】editor's decisive concerns and the work done for each ...
【New material】recoding rounds / conditions / theory rebuild — status ...
【Letter】summary page + located point-by-point complete? ...
【Sweep】terminology, exhibits, declarations, anonymity re-checked? ...
【Next step】aos-submission (final preflight) → resubmit in Editorial Manager
Version History
- 9f86f09 Current 2026-07-19 14:23
