reflect-paper

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Reflect Paper

论文反思

Run a structured reflection loop over a research-paper draft. Treat this as two separate perspectives, then merge them into one revision plan:

Researcher pass — technical truth, evidence, citations, venue fit, novelty, limitations, reviewer risk.
Writer pass — narrative, framing, section structure, abstract, related work placement, prose, figures/tables, reader flow.

This is an orchestration skill: it calls on sibling skills when their narrower checks are needed, but it owns the reflection protocol and final synthesis.

If the runtime supports delegated agents or subagents, split the reflection into two independent workers:

Send a Researcher worker the draft, target venue/context, and the Researcher pass checklist only.
Send a Writer worker the draft, target venue/context, and the Writer pass checklist only.
Merge both outputs yourself; do not let either worker decide the final verdict alone.

If delegation is unavailable, run the same passes sequentially. Keep separate notes for each pass so technical-risk findings do not flatten writing feedback, and writing polish does not hide correctness concerns.

对研究论文草稿进行结构化的反思循环。将其视为两个独立的视角，然后合并为一份修订计划：

Researcher环节 — 技术真实性、证据、引用、适配的会议/期刊、创新性、局限性、审稿风险。
Writer环节 — 叙事、框架、章节结构、摘要、相关工作的布局、措辞、图表、读者阅读流畅度。

这是一个编排类skill：当需要更细分的检查时，它会调用同级skill，但它主导反思流程并负责最终的整合。

如果运行环境支持委托代理或子代理，将反思拆分为两个独立的工作单元：

向Researcher工作单元发送草稿、目标会议/期刊背景以及仅包含Researcher环节的检查清单。
向Writer工作单元发送草稿、目标会议/期刊背景以及仅包含Writer环节的检查清单。
自行合并两个输出结果；不要让任何一个工作单元单独决定最终结论。

如果不支持委托，则依次执行相同的环节。为每个环节保留单独的记录，这样技术风险发现不会掩盖写作反馈，写作润色也不会隐藏正确性问题。

When to use

使用场景

"Reflect on this paper before I submit"
"Run a research agent and writer agent on my draft"
"Critique this draft from both technical and writing angles"
"Give me a revision plan"
"What should I fix before sending this to my advisor / reviewers?"
"Do a final paper reflection pass"

"提交前帮我反思这篇论文"
"在我的草稿上运行research agent和writer agent"
"从技术和写作角度批评这份草稿"
"给我一份修订计划"
"提交给导师/审稿人前我应该修改哪些内容？"
"进行最终的论文反思环节"

Inputs

输入内容

The draft: LaTeX source, PDF, markdown, Word-exported text, or pasted sections.
Optional: target venue or CFP URL.
Optional: bibliography file, related-work notes, reviews, or advisor feedback.
Optional: the user's current goal, such as submission readiness, camera-ready polish, rebuttal, or presentation prep.

草稿：LaTeX源码、PDF、markdown、Word导出文本，或粘贴的章节内容。
可选：目标会议/期刊或CFP（Call for Papers）URL。
可选：参考文献文件、相关工作笔记、评审意见或导师反馈。
可选：用户当前目标，比如提交准备、终稿润色、反驳准备或演讲准备。

Process

流程

1. Scope The Reflection

1. 确定反思范围

Identify the lifecycle stage and what "good" means for this pass:

Stage	Primary concern
Early draft	Missing contribution, weak framing, incomplete evidence.
Pre-submission	Venue fit, page budget, anonymization, citation reliability, reviewer risk.
Rebuttal	Review coverage, evidence anchors, tone, response budget.
Camera-ready	Required final-file steps, de-anonymization, acknowledgments, source consistency.
Presentation	Talk story, timing, Q&A risk, slide/poster readability.

If a venue or CFP is involved, use

parse-cfp

add-venue-profile

before relying on page limits, blind level, required sections, or deadlines.

确定论文的生命周期阶段以及本次反思的"合格"标准：

阶段	主要关注点
早期草稿	缺失的贡献、薄弱的框架、不完整的证据。
提交前	适配的会议/期刊、页数限制、匿名化、引用可靠性、审稿风险。
反驳阶段	评审覆盖范围、证据锚点、语气、回应篇幅限制。
终稿阶段	最终文件的必要步骤、去匿名化、致谢、源码一致性。
演讲准备	演讲叙事、时长、问答风险、幻灯片/海报可读性。

如果涉及会议/期刊或CFP，在依赖页数限制、匿名级别、必填章节或截止日期之前，先使用

parse-cfp

或

add-venue-profile

。

2. Researcher Pass

2. Researcher环节

Read for correctness and evidence. Check:

Core claim: what the paper says it contributes, and whether the evidence actually supports it.
Novelty and related work: missing comparisons, closest competitors, thin or overstated deltas.
Citations: unresolved entries, suspicious metadata, fabricated-looking references, uncited claims.
Methods and evaluation: assumptions, baselines, ablations, datasets, reproducibility, statistics, limitations.
Venue fit: track expectations, page budget, anonymity, required sections, LLM/AI-use policy, artifact expectations.
Reviewer risk: likely objections and what evidence would defuse them.

Delegate as needed:

Need	Use
Find missing related work	`find-papers` , `literature-review`
Fetch legal OA copies	`fetch-paper`
Check bibliography	`verify-citations`
Study venue exemplars	`study-exemplars`
Simulate reviewer risk	`simulate-reviewers`
Check CFP/venue facts	`parse-cfp` , `select-venue` , `tailor-to-venue`

Record findings as:

finding

evidence

risk

suggested fix

, and

confidence

检查正确性和证据。需检查：

核心主张：论文声称的贡献，以及证据是否真正支持该主张。
创新性与相关工作：缺失的对比、最接近的竞品、薄弱或夸大的差异。
引用：未解决的条目、可疑的元数据、看似伪造的参考文献、无引用的主张。
方法与评估：假设、基线、消融实验、数据集、可复现性、统计数据、局限性。
适配的会议/期刊：领域预期、页数限制、匿名性、必填章节、LLM/AI使用政策、成果要求。
审稿风险：可能的异议以及能化解这些异议的证据。

按需委托：

需求	使用的skill
查找缺失的相关工作	`find-papers` , `literature-review`
获取合法的开放获取版本	`fetch-paper`
检查参考文献	`verify-citations`
研究会议/期刊范例	`study-exemplars`
模拟审稿风险	`simulate-reviewers`
检查CFP/会议/期刊信息	`parse-cfp` , `select-venue` , `tailor-to-venue`

记录发现内容：

发现

、

证据

、

风险

、

建议修正方案

和

置信度

。

3. Writer Pass

3. Writer环节

Read for communication and persuasion. Check:

One-sentence paper promise: can a reader say what the paper does and why it matters?
Abstract: motivation, gap, approach, result, impact, and venue-specific length/keyword rules.
Introduction: problem setup, contribution bullets, reader contract.
Related work: clustered themes and explicit deltas, not a laundry list.
Method and evaluation: section order, figure/table references, claim titles, readable transitions.
Prose: overclaiming, hedging, AI-sounding filler, terminology drift, passive contribution statements.
Figures and tables: captions, crossrefs, color accessibility, layout, and whether each visual answers a reader question.

Delegate as needed:

Need	Use
Abstract rewrite or lint	`write-abstract`
Related work drafting	`draft-related-work`
Full review synthesis	`literature-review`
Prose polish	`polish-prose`
Tables and figures	`polish-tables-figures`
Talk/poster translation	`make-slides` , `write-talk-script` , `make-poster`

Record findings as:

reader problem

location

revision

expected benefit

, and

next skill

检查沟通效果与说服力。需检查：

一句话论文承诺：读者能否说出论文做了什么以及为什么重要？
摘要：动机、研究缺口、方法、结果、影响，以及符合会议/期刊要求的长度/关键词规则。
引言：问题设定、贡献要点、读者约定。
相关工作：按主题分类并明确差异，而非简单罗列。
方法与评估：章节顺序、图表引用、主张标题、易读的过渡。
措辞：过度主张、含糊其辞、AI生成的冗余内容、术语不一致、被动语态的贡献陈述。
图表：标题、交叉引用、色彩可访问性、布局，以及每个视觉元素是否能解答读者的问题。

按需委托：

需求	使用的skill
摘要重写或检查	`write-abstract`
相关工作撰写	`draft-related-work`
完整评审整合	`literature-review`
措辞润色	`polish-prose`
表格与图表优化	`polish-tables-figures`
转换为演讲/海报	`make-slides` , `write-talk-script` , `make-poster`

记录发现内容：

读者问题

、

位置

、

修订内容

、

预期收益

和

下一步skill

。

4. Reflection Merge

4. 反思结果整合

Merge the two passes into a prioritized plan. Do not average away conflicts: when the Researcher pass wants more caveats and the Writer pass wants a sharper claim, state the tradeoff and propose a wording that is both accurate and readable.

Use this priority scale:

Priority	Meaning
P0	Blocks submission or creates serious correctness/integrity risk.
P1	Likely affects review outcome or reader understanding.
P2	Improves clarity, polish, or confidence.
P3	Nice-to-have cleanup.

将两个环节的结果合并为一份优先级明确的计划。不要平均化冲突：当Researcher环节需要更多说明而Writer环节需要更明确的主张时，说明权衡并提出既准确又易读的措辞。

使用以下优先级量表：

优先级	含义
P0	阻碍提交或造成严重的正确性/诚信风险。
P1	可能影响评审结果或读者理解。
P2	提升清晰度、润色效果或可信度。
P3	锦上添花的整理工作。

5. Output

5. 输出结果

Return a concise reflection report:

Verdict — ready / close / needs major revision / not ready.
Researcher pass — top technical, evidence, citation, and venue risks.
Writer pass — top framing, structure, prose, and visual risks.
Merged action plan — prioritized table with owner/pass, location, action, next skill, and estimated effort.
Open questions — facts that require user input or live verification.

返回一份简洁的反思报告：

结论 — 已就绪/接近就绪/需要重大修订/未就绪。
Researcher环节 — 最主要的技术、证据、引用和会议/期刊风险。
Writer环节 — 最主要的框架、结构、措辞和视觉呈现风险。
合并行动计划 — 按优先级排序的表格，包含负责人/环节、位置、行动、下一步skill和预估工作量。
未解决问题 — 需要用户输入或实时验证的信息。

Guardrails

约束规则

Never invent papers, results, deadlines, page limits, or venue rules.
Do not modify technical claims to sound stronger unless the evidence supports them.
Keep confidential reviews and unpublished paper content local.
For legal/copyright safety, fetch papers only from legal open-access sources and do not reproduce abstracts or paper text in the report.

切勿虚构论文、结果、截止日期、页数限制或会议/期刊规则。
除非证据支持，否则不要修改技术主张使其听起来更有说服力。
保密评审意见和未发表的论文内容。
为了法律/版权安全，仅从合法的开放获取来源获取论文，且不要在报告中复制摘要或论文文本。