seven-pass-review

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Seven-Pass Adversarial Review

七轮对抗式评审

Runs seven independent reviewers, each focused on a single lens, then synthesizes their findings into one prioritized revision plan. Pattern 15 from the workflow guide, mechanized.

Why seven passes? A single-agent review blends lenses and softens each one. Seven forked agents each approach the paper with full context budget for their own lens, then a synthesizer resolves conflicts and de-duplicates.

When to pick this over
/review-paper
: This skill costs roughly 7× more tokens than
/review-paper
(default) and ~2× more than
/review-paper --adversarial
. Use it when the paper is submission-ready or at R&R stage and you need maximum lens coverage. For early drafts or iterative work,
/review-paper
is the right tool. For journal-simulation pressure test, use
/review-paper --peer <journal>
instead.

启动七个独立评审Agent，每个Agent聚焦单一评审维度，随后将它们的评审结果整合为一份按优先级排序的修订方案。这是工作流指南中的模式15的机械化实现。

为什么要七轮评审？ 单Agent评审会混合不同维度的视角，弱化每个维度的评审力度。七个分支Agent各自拥有充足的上下文预算专注于自身的评审维度，之后由一个合成Agent解决冲突并去重。

何时选择该工具而非
/review-paper
：该技能的token消耗约为默认版
/review-paper
的7倍，约为
/review-paper --adversarial
的2倍。当论文已准备好提交或处于修改再审阶段，且需要全面覆盖所有评审维度时使用。对于早期草稿或迭代工作，
/review-paper
是更合适的工具。如需模拟期刊压力测试，请使用
/review-paper --peer <journal>
。

Inputs

输入项

```
$0
```
— manuscript path (
```
.tex
```
,
```
.qmd
```
,
```
.md
```
, or
```
.pdf
```
). Required.

```
$0
```
— 手稿路径（
```
.tex
```
、
```
.qmd
```
、
```
.md
```
或
```
.pdf
```
格式）。必填项。

The Seven Lenses

七个评审维度

Each lens runs as a forked subagent (context: fork) so the main conversation stays clean.

#	Lens	Focus	Agent type
1	Abstract audit	Does the abstract state the question, method, result, and contribution? Does it match the paper?	general-purpose
2	Intro structure	Does the intro follow Cochrane / Varian framework? Literature placement? Contribution clarity?	general-purpose
3	Methods / identification	Are assumptions stated? Is identification credible? Are alternatives addressed?	domain-reviewer
4	Results + tables	Do tables read standalone? Is magnitude + significance discussed? Units consistent?	general-purpose
5	Robustness	Are obvious threats pre-empted? Is the robustness section convincing or theatrical?	general-purpose
6	Prose quality	Sentence-level clarity, hedging, passive voice, paragraph cohesion	proofreader
7	Citation audit	Invokes `/validate-bib --semantic` ; checks cite-claim direction for top-10 works	general-purpose

每个维度由一个分支子Agent（上下文：fork）运行，确保主对话保持整洁。

序号	评审维度	聚焦点	Agent类型
1	摘要审核	摘要是否明确阐述了研究问题、方法、结果和贡献？是否与论文内容一致？	general-purpose
2	引言结构	引言是否遵循Cochrane/Varian框架？文献综述的位置是否恰当？贡献是否清晰？	general-purpose
3	方法/识别性	是否明确说明假设？识别性是否可信？是否讨论了替代方案？	domain-reviewer
4	结果与表格	表格是否可独立阅读？是否讨论了结果的量级和显著性？单位是否一致？	general-purpose
5	稳健性	是否预先应对了明显的质疑？稳健性部分是否具有说服力还是流于形式？	general-purpose
6	文风质量	句子层面的清晰度、措辞严谨性、被动语态使用、段落连贯性	proofreader
7	参考文献审核	调用 `/validate-bib --semantic` ；检查引用量前十的文献的引用主张方向是否正确	general-purpose

Workflow

工作流

Phase 0: Pre-flight

阶段0：预检查

Resolve manuscript path.
Decide if
```
.pdf
```
→ extract text first (
```
pdftotext -layout
```
).
Create output dir:
```
quality_reports/seven_pass_[stem]/
```
.

解析手稿路径。
若为
```
.pdf
```
格式，先提取文本（使用
```
pdftotext -layout
```
）。
创建输出目录：
```
quality_reports/seven_pass_[stem]/
```
。

Phase 1: Spawn 7 reviewers in parallel

阶段1：并行生成7个评审Agent

In a single message, spawn 7 Task tool calls (one per lens). Each subagent gets:

The manuscript path (to re-read with its own context).
The lens-specific prompt (below).

Instructions to write to

quality_reports/seven_pass_[stem]/lens_[N]_[lens-name].md

Severity tagging: CRITICAL / MAJOR / MINOR.

Lens prompt rubrics are embedded inline below — one summary paragraph per lens. Each forked subagent receives its lens's rubric plus the manuscript path.

Lens prompt summaries:

Lens 1 (Abstract): Does the first sentence state the question? Does it name the method? Quantify the headline result? State one-sentence contribution? Cross-check: do these four things match the body?
Lens 2 (Intro): Does the intro open with the question? Hook → context → contribution → roadmap? Lit review placed correctly (after the hook, not before)? Contribution-counted (1, 2, 3…)? Preview of findings with magnitudes?
Lens 3 (Methods): Is every assumption stated? Are they strong or weak? Is identification one-liner clear? Are known violations (selection, measurement, reverse causality, SUTVA) addressed? Are instruments / RDD / DiD assumptions explicit and defensible?
Lens 4 (Results): Does each table read standalone (caption, units, SEs clarified)? Is magnitude interpreted (not just significance)? Are units consistent across tables? Are figures legible at 8pt?
Lens 5 (Robustness): Does the paper ANTICIPATE a sharp referee's objections? Are robustness checks motivated, or just listed? Power/placebo tests present? Heterogeneity explored where promised?
Lens 6 (Prose): Sentences under 30 words? Active voice dominant? Hedging proportionate (neither overclaiming nor endless "may suggest")? Paragraph topic sentences?
Lens 7 (Citations): Invoke
```
/validate-bib --semantic
```
. For top-10 cited works, does the in-text claim match the cited paper's actual finding direction? Are contemporary / competing works cited?

在一条消息中，生成7个Task工具调用（每个维度对应一个）。每个子Agent将获得：

手稿路径（用于以自身上下文重新读取内容）。
对应维度的提示词（如下所示）。

写入

quality_reports/seven_pass_[stem]/lens_[N]_[lens-name].md

的指令。

严重程度标签：CRITICAL / MAJOR / MINOR。

维度提示词规则内嵌在下方——每个维度对应一段总结性文字。每个分支子Agent将收到对应维度的规则及手稿路径。

维度提示词摘要：

维度1（摘要）： 第一句是否明确研究问题？是否提及方法？是否量化核心结果？是否用一句话说明贡献？交叉验证：这四点是否与正文内容一致？
维度2（引言）： 引言是否以研究问题开篇？是否遵循“钩子→背景→贡献→路线图”的结构？文献综述位置是否正确（在钩子之后而非之前）？是否明确列出贡献点（1、2、3…）？是否预览了带量级的研究结果？
维度3（方法）： 是否明确说明所有假设？假设是强假设还是弱假设？识别性的核心逻辑是否清晰？是否讨论了已知的违反情况（选择偏差、测量偏差、反向因果、SUTVA）？工具变量/断点回归/双重差分的假设是否明确且具有可辩护性？
维度4（结果）： 每个表格是否可独立阅读（标题、单位、标准误是否清晰）？是否解读了结果量级（而非仅讨论显著性）？各表格间单位是否一致？图表在8pt字号下是否清晰可读？
维度5（稳健性）： 论文是否预判了严苛审稿人的质疑？稳健性检验是否有依据，还是仅简单罗列？是否包含功效/安慰剂检验？是否按预期探索了异质性？
维度6（文风）： 句子是否控制在30词以内？是否以主动语态为主？措辞严谨性是否适度（既不过度断言也不滥用“可能表明”）？段落是否有主题句？
维度7（参考文献）： 调用
```
/validate-bib --semantic
```
。对于引用量前十的文献，正文引用的主张是否与被引论文的实际结论方向一致？是否引用了当代/竞争研究成果？

Phase 2: Synthesize

阶段2：结果合成

Wait for all 7 lens reports. Then read them and produce:

quality_reports/seven_pass_[stem]/_SYNTHESIS.md

markdown

undefined

等待所有7个维度的评审报告完成。随后读取这些报告并生成：

quality_reports/seven_pass_[stem]/_SYNTHESIS.md

markdown

undefined

Seven-Pass Review: [Manuscript]

七轮评审：[手稿名称]

Date: YYYY-MM-DD Path: [manuscript]

日期： YYYY-MM-DD 路径： [手稿路径]

Executive verdict

执行 verdict

Overall state: [SUBMIT / REVISE-MINOR / REVISE-MAJOR / REJECT-AND-RESTART]

整体状态： [SUBMIT / REVISE-MINOR / REVISE-MAJOR / REJECT-AND-RESTART]

Cross-lens CRITICAL issues

跨维度CRITICAL问题

#	Lens(es)	Issue	Recommendation

序号	涉及维度	问题	建议

MAJOR issues (second-round)

MAJOR问题（第二轮）

#	Lens(es)	Issue

序号	涉及维度	问题

MINOR polish

MINOR优化点

[bulleted]

[项目符号列表]

Per-lens scorecard

各维度评分卡

Lens	Critical	Major	Minor	Score/10
1. Abstract
2. Intro
3. Methods
4. Results
5. Robustness
6. Prose
7. Citations
Overall

维度	Critical	Major	Minor	得分/10
1. 摘要
2. 引言
3. 方法
4. 结果
5. 稳健性
6. 文风
7. 参考文献
整体

Revision plan (in recommended order)

修订计划（推荐顺序）

[Highest-leverage fix — usually a lens with 2+ CRITICALs]
…
[Lowest-leverage polish]

[最高优先级修复——通常是存在2个及以上CRITICAL问题的维度]
…
[最低优先级优化]

Contradictions between lenses

维度间矛盾点

[If two lenses disagree, surface here. E.g., Lens 2 says "expand contribution" but Lens 6 says "trim intro".]

undefined

[若两个维度的评审意见不一致，在此列出。例如：维度2建议“扩展贡献阐述”，但维度6建议“精简引言”。]

undefined

Phase 3: Token-budget report

阶段3：Token预算报告

After synthesis, print:

Seven-pass review complete.
Subagents: 7 (parallel) + 1 synthesizer.
Approx token usage: ~80–120k (vs ~15k for single-pass /review-paper).
Runtime: ~3–5 min wall-clock.
For cheaper alternatives:
  - Single-pass: /review-paper
  - Iterative: /review-paper --adversarial

合成完成后，输出：

七轮评审完成。
子Agent数量：7个（并行） + 1个合成Agent。
预估Token使用量：约80–120k（对比单轮`/review-paper`的约15k）。
运行时间：约3–5分钟。
如需更经济的替代方案：
  - 单轮评审：/review-paper
  - 迭代评审：/review-paper --adversarial

When to use this skill

何时使用该技能

Before first submission to a top journal.
After a major revision when you want to catch drift.
R&R when referees disagree — surfaces contradictions your revision must navigate.

首次提交顶级期刊前。
重大修订后，用于检查内容偏离情况。
修改再审阶段审稿人意见不一致时——可梳理出修订需解决的矛盾点。

When NOT to use

何时不使用

Early drafts (use
```
/review-paper
```
single-pass first).
Short notes, comments, or replies (overkill).
When you've already run this in the last 7 days and nothing substantive changed.

早期草稿（先使用单轮
```
/review-paper
```
）。
短篇笔记、评论或回复（过于冗余）。
过去7天内已运行过该技能且手稿未发生实质性修改时。

Cross-references

交叉引用

.claude/skills/review-paper/SKILL.md

— the single-pass and

--adversarial

modes (cheaper, faster).

```
.claude/skills/validate-bib/SKILL.md
```
— invoked by Lens 7.
```
.claude/skills/audit-reproducibility/SKILL.md
```
— complementary; numeric-claims side of the audit.
Workflow guide, Pattern 15 — the narrative explanation of why seven lenses.

.claude/skills/review-paper/SKILL.md

— 单轮及

--adversarial

模式（更经济、更快）。

```
.claude/skills/validate-bib/SKILL.md
```
— 维度7调用的工具。
```
.claude/skills/audit-reproducibility/SKILL.md
```
— 补充工具；用于审核数值主张。
工作流指南，模式15 — 关于七个评审维度必要性的说明。

Exit behavior

退出行为

Exits 0 always (review is informational). The synthesis report's "Executive verdict" is the gate.
Any
```
CRITICAL
```
at the top of the synthesis should block submission until resolved.

始终以0状态退出（评审仅提供信息）。合成报告中的“执行 verdict”是决策依据。
合成报告顶部的任何
```
CRITICAL
```
问题应在解决前阻止提交。

What this skill does NOT do

该技能不具备的功能

Re-run seven lenses if the manuscript hasn't changed — check git diff against last run date in
```
_SYNTHESIS.md
```
, skip unchanged lenses if requested via
```
--incremental
```
(future).
Auto-apply fixes — that's
```
/review-paper --adversarial
```
's job.
Replace human judgment. A reviewer who knows your subfield still beats seven LLMs.

若手稿未发生变化，不会重新运行七轮评审——对比
```
_SYNTHESIS.md
```
中的上次运行日期与git diff，若通过
```
--incremental
```
请求，将跳过未修改的维度（未来功能）。
不会自动应用修改——这是
```
/review-paper --adversarial
```
的功能。
无法替代人工判断。了解你所在细分领域的审稿人仍优于七个LLM。