zrwangems/dsh-zrwang-plugin / ars/skills/academic-paper-reviewer
academic-paper-reviewer
Flagged模拟同行评审:期刊匹配审稿人 + 3 位动态领域审稿人 + 魔鬼代言人,输出编辑决定(接收/小修/大修/拒稿)与返修路线图,可做方法论专项评审与再审核验。Triggers: review paper, peer review, manuscript review, check methodology, re-review | 审稿, 同行评审, 评审论文, 检查方法, 复审。只做评审,不改写稿件。
references/review_quality_thinking.md
View on GitHub> 输出语言(本 DSH 集成强制):所有面向用户的文字、稿件、摘要、评审与报告一律使用**简体中文**;英文术语与引文保持英文。文中的 `zh-TW` / `zh-tw-en` 只是上游语言代码与脚本参数,不代表要用繁体输出。用户明确要求其他语言时以用户为准。
>
> 多角色执行(评审团、十阶段流水线、角色扮演)必须使用 DSH 的 Agent Team 工具(`spawn_teammate`、`team_task_*`、`wait_agent`、`send_message`);当前会话看不到这些工具,说明用户没开 Agent Team,必须先提示用户开启,不得退化成单会话顺序扮演。
# Review Quality Thinking Framework
A cognitive framework for producing high-quality reviews. Teaches **how to think** about paper quality, not just which judgement label to emit.
## The Three Lenses
Every paper should be evaluated through three lenses simultaneously:
### Lens 1: Internal Validity — "Does the evidence support the claims?"
Ask in order:
1. What is the central claim?
2. What evidence is presented?
3. Is there a logical chain from evidence to claim? (Warrant)
4. Are there alternative explanations the authors didn't consider?
5. Would removing any single piece of evidence collapse the argument?
**If #5 is yes**: The argument depends on a single linchpin. Flag it — the paper's contribution is only as strong as that one piece of evidence.
### Lens 2: External Validity — "Does this matter beyond this study?"
Ask in order:
1. Who is the population of interest?
2. Does the sample represent that population?
3. Are the conditions replicable?
4. Would the findings hold in a different context/culture/time?
5. What are the boundary conditions the authors don't mention?
**Judgment heuristic**: Check the manuscript's claimed scope of generalization against its sampling frame — a criterion-only comparison, never an author-population base rate (#574 B1). If the sample is from one university in one country, findings cannot claim to apply "generally" without qualification; if the claims already match the evidence's scope, there is no finding here.
### Lens 3: Contribution — "So what?"
Ask in order:
1. What did we know before this paper?
2. What do we know after?
3. Is the delta meaningful (not just statistically significant)?
4. Who benefits from knowing this?
5. What new questions does this open?
**Judgment heuristic**: If you can't articulate the delta in one sentence, either the contribution is weak or the paper hasn't communicated it clearly. Both are review-worthy observations.
## Common Reviewer Traps
| Trap | Description | How to Avoid |
|------|-------------|-------------|
| **Methodological tunnel vision** | Only critiquing methods, ignoring whether the question matters | Start with Lens 3 (contribution) before Lens 1 |
| **Novelty bias** | Penalizing replication or incremental work | Replication IS valuable; evaluate on execution quality |
| **Expertise projection** | Expecting the paper to use your preferred method | Evaluate the chosen method on its own terms |
| **Positivity-severity oscillation** | Being too nice in comments, too harsh in the final judgement | Commit the criterion, evidence, and categorical judgement before polishing the comments |
| **Missing forest for trees** | Listing 20 minor issues, missing the one fatal flaw | Always state the single most important issue first |
## Calibration Questions (ask after drafting your review)
1. If this paper were published as-is, would it mislead readers? (If yes → Major Revision or Reject)
2. Could the authors reasonably address my concerns in one revision cycle? (If no → Reject)
3. Am I being harder on this paper than I would be on my own work? (Calibration check)
4. Did I acknowledge every genuine strength I found — and manufacture none? (Evidence-driven balance check, #574 A1/B1)
5. Would my review help the authors improve, even if the paper is rejected? (Constructiveness check)
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |