科研 Skill
人文手稿整体评审
reviewing-manuscript
Performs a whole-document review of a manuscript’s question, scope, argument, citations, and structure. Use for a holistic quality check before revision.
SKILL.md
Reviewing a Manuscript
Purpose
Review a manuscript as an integrated scholarly argument rather than as a collection of polished paragraphs. The review identifies strengths, blocking weaknesses, revision priorities, and decisions that remain the researcher's responsibility.
This skill is not a simulated hostile reviewer and not a guarantee of acceptance. Its task is to reconstruct the manuscript fairly, test whether its parts support one another, and produce an actionable revision map.
Contract
Accepts a complete or substantial manuscript, research brief, argument map, journal guidelines, prior gate reports, citation audit, terminology ledger, or reviewer criteria.
Requires enough of the manuscript to reconstruct its central question, thesis, evidence, method, and contribution. A target genre or venue is strongly recommended.
Produces a manuscript review, reconstructed argument, strengths, findings by severity, revision priorities, venue-fit observations where evidence is available, and a manuscript-review gate report.
May produce section-level annotations, a revision sequence, an editorial triage table, or handoffs to citation, terminology, scope, argument, or peer-review-response skills.
Fails when the manuscript is unavailable, only a title or abstract is supplied for whole-manuscript judgment, the requested review requires invented sources or reviewer opinions, or venue criteria cannot be verified.
Guarantees that major criticisms are linked to observable manuscript evidence and that severity is distinguished from stylistic preference.
Does not guarantee publication, reviewer agreement, originality, factual truth, or disciplinary adequacy beyond the materials and expertise available.
When to use
Use after a coherent draft exists, before external circulation, after major structural revision, before selecting final submission status, or when a researcher needs a prioritized diagnosis rather than line editing.
Do not use this skill as a substitute for citation audit, specialist fact checking, or copyediting. It may identify the need for those operations and route the manuscript accordingly.
Inputs required
Provide:
- The manuscript, including notes and bibliography where possible.
- The intended article, chapter, conference, or book genre.
- Target venue guidelines or evaluation criteria, if fit is requested.
- The researcher's stated question, thesis, and claimed contribution.
- Available argument maps, audit reports, or reviewer history.
- Any sections that must not be rewritten without approval.
If the stated thesis differs from what the manuscript actually argues, preserve both and report the discrepancy.
Procedure
1. Establish review scope
State what was received, what was not received, which criteria will be used, and whether citation content, factual claims, language quality, and venue fit are included or only flagged for separate review.
2. Reconstruct the manuscript charitably
Before criticizing, summarize:
- research problem;
- central question;
- thesis;
- method or interpretive practice;
- corpus or evidence;
- chapter or section sequence;
- claimed contribution.
Use the manuscript's strongest defensible formulation. If reconstruction requires guessing, mark the ambiguity.
3. Test question–thesis alignment
Determine whether the thesis answers the research question and whether the conclusion answers the same question introduced at the beginning. Flag a moving research object, broadened conclusion, or question that is descriptive while the thesis is causal or normative.
4. Test scope and evidence coverage
Compare the manuscript's claims with its corpus, historical range, linguistic range, and method. Identify where local readings become field-wide, historical, technological, social, or universal claims without sufficient bridge evidence.
5. Test concept architecture
Check whether central concepts are defined, distinguished from adjacent terms, used at a stable level of analysis, and connected to their sources or lineage. A theoretically impressive vocabulary is not a contribution if it does no identifiable analytical work.
6. Test argument structure
Reconstruct major claims, evidence, warrants, objections, and consequences. Identify missing warrants, circular reasoning, unsupported transitions, redundant sections, late definitions, and conclusions introduced before evidence.
Ask whether each section changes what the reader can reasonably accept. Sections that only repeat context or theorist summaries may need compression or relocation.
7. Test close readings and examples
Assess whether observed features are distinguished from interpretation, whether alternative readings are addressed, and whether examples support the scale of the claim. Do not judge an interpretation solely because another reading is possible; judge whether the manuscript makes its interpretive path accountable.
8. Test literature positioning and contribution
Determine whether prior work is represented as a structured scholarly conversation rather than a list. Evaluate whether the contribution is stated as a bounded intervention: new object, relation, distinction, reinterpretation, method, synthesis, or consequence.
Do not accept “no one has studied this” without documented review scope.
9. Inspect citation and terminology status
Use existing audit reports if provided. Otherwise flag risks without claiming full verification. Identify central claims whose support is unclear, direct quotations needing audit, inconsistent key terms, and translations requiring specialist confirmation.
10. Classify findings by severity
Use:
- Fatal: the central thesis cannot currently be sustained.
- Major: substantial new evidence, reframing, or reorganization is required.
- Moderate: the core can stand, but an important section or distinction needs revision.
- Minor: localized clarity, style, or formatting issue.
Severity MUST reflect argumentative consequence, not reviewer irritation.
11. Prioritize revision
Organize recommendations in dependency order. Do not begin with prose polishing when the research question, evidence, or argument structure may change. A typical sequence is:
question/scope -> concept -> evidence -> argument -> objections -> citations -> terminology -> prose/format
For each major issue state the observed evidence, why it matters, the smallest adequate remedy, and which skill should handle it.
12. Issue the gate report
Return PASS only when no blocking issue remains within the declared review scope. Use CONDITIONAL PASS when the argument is viable but external audit, specialist review, or localized revision remains. Return FAIL when central evidence, scope, concept, or logic does not support the thesis.
Stop signals
Stop or narrow the review when:
- only an abstract is available for a whole-manuscript judgment;
- source verification is requested without sources;
- journal instructions cannot be verified;
- the review would require specialist linguistic, archival, legal, or statistical competence not available;
- the requester asks for praise, hostility, or a predetermined accept/reject verdict;
- anonymous reviewer identities or motives are requested without evidence.
Completion criteria
A review is complete when:
- the manuscript is fairly reconstructed;
- every major criticism cites an observable feature or location;
- question, scope, concepts, evidence, argument, interpretation, contribution, citations, and terminology have been considered;
- findings are severity-ranked;
- revision actions are dependency-ordered;
- external-review needs are explicit;
- the gate status follows from the findings.
Anti-fabrication rules
Do not invent reviewer consensus, journal preferences, source contents, missing sections, disciplinary trends, or likely acceptance rates. Do not claim a manuscript is original because no duplicate wording was noticed. Do not attribute motives to the author. Do not manufacture criticisms merely to appear rigorous.
When evidence is unavailable, write “not assessed” or route to the appropriate audit.
Output format
Provide:
- Review scope and limitations.
- Charitable reconstruction.
- Principal strengths.
- Findings by severity, each with evidence and consequence.
- Revision priorities in dependency order.
- Required specialist or external checks.
- Recommended next skills.
Gate report
Gate: Manuscript Review
Status: PASS | CONDITIONAL PASS | FAIL
Evidence checked:
Blocking issues:
Non-blocking risks:
Researcher decisions required:
Required next action:
Next skill:
Good invocation
Review this article as an integrated humanities argument. Reconstruct it fairly, rank problems by severity, cite section locations, and do not claim that citations are verified unless an audit report supports that claim.
Bad invocation
Pretend to be an angry Reviewer 2, invent devastating objections, and tell me the journal will reject it.
The bad invocation requests theater and unsupported prediction rather than evidence-based review.
Next skills
stress-testing-argumentfor major objections.auditing-citationsfor evidentiary verification.checking-terminology-consistencyfor concept drift.responding-to-peer-reviewwhen real reviewer comments exist.verifying-before-submissionafter revision.
Friction triggers
Create structural-coherence friction for hidden contradictions, invalidated gates, reopened decisions, unresolved evidence conflicts, or fluency that masks argumentative gaps.
Friction checks
Report open friction and separate rhetorical fluency from argument coherence. Inspect whether counterevidence, rival readings, and decision history remain visible.
Judgment boundary
Review recommendations do not become author decisions. Rank options and consequences, then request researcher action where revision changes thesis, scope, or interpretation.
Evidence ledger updates
Reference ledger gaps and contradictory standing for each affected claim; do not infer support from citation density or polished discussion.
Interpretation history impact
Review the active version and its supersession chain. A new review interpretation remains proposed until researcher response.
Possible epistemic return
New review evidence may reopen question, scope, concept, interpretation, or argument. Known defects discovered in review use rollback.
Productive refusal
If the supplied manuscript cannot support a whole-document judgment, state inspected range, unavailable artifacts, bounded findings, and required review inputs.
Gate impact
Summarize invalidated and unrechecked gates. Whole-manuscript PASS is prohibited while a blocking friction or controlling invalid validity remains.
Limitations
A manuscript review is conditional on disciplinary context, source access, and the declared criteria. It cannot replace expert peer review, editor judgment, language editing, legal review, ethics review, or empirical validation. Different qualified reviewers may disagree; the skill makes reasons and dependencies explicit rather than claiming a single final verdict.
示例提示词
帮我对这份手稿做全文评审,检查研究问题、论证和引用是否站得住脚。
所属插件
安装该插件会一并获得同领域的 62 个 Skill / MCP。
同插件的其他科研 Skill
- 学术文献获取下载Auto-downloads open-access PDFs for graded sources and writes a manual worklist for paywalled items. Use after literature-review, before ingest-source.
- 单篇文献入库管理Adds one verified PDF to a shared bibliography library, computing its citation key and keywords. Use to register a single document outside the full workflow.
- 数字人文计算方法Surveys computational DH methods: text mining, NLP, corpus linguistics, archive standards, GIS, network analysis, and OCR. Use for an overview of available DH tools.
- 考古地理信息系统分析Runs archaeological GIS analysis with GeoPandas: catchment analysis, viewshed, kernel density, spatial statistics. Use for mapping archaeological site locations.
- 论证结构提取可视化Extracts and visualizes argument structure: Toulmin reconstruction, support/attack graphs, circularity detection. Use to diagram an argument’s logical structure.
- 考古遗物类型学分析Performs artifact analysis: typological classification, ceramic attribute coding, lithic reduction analysis. Use when classifying excavated artifacts statistically.
- 引文与出处核查Checks citations, quotations, page numbers, and bibliography entries for accuracy and source fit. Use before submission to verify source-supported claims.
- 研究选题构思设计Explores research intent, field state, methodology, and ethics, then writes an approved design doc. Use before starting a new research question or chapter.
- 术语译名一致性检查Audits a manuscript for terms, translations, abbreviations, and names whose forms must stay stable. Use when terminology drift could undermine an argument.
- 计算语用学分析Runs computational pragmatics: speech act classification, implicature detection, discourse coherence, politeness analysis. Use for language-use-and-meaning studies.
- 文献综述学术对话构建Turns verified sources into a structured scholarly conversation among positions, not a chronological summary. Use when synthesizing literature into a debate narrative.
- 语料库语言学分析Performs corpus analysis with NLTK/spaCy: collocation, keyword analysis, KWIC concordance, word2vec, diachronic trends. Use for quantitative text-corpus research.
- 数字档案馆资源检索Queries digital archive REST APIs: Europeana, Chronicling America, Internet Archive, HathiTrust, DPLA, OAI-PMH. Use when searching or harvesting heritage metadata.
- 学术手稿章节撰写Drafts a chapter or article section by section from synthesized wiki content, agreeing each section’s argument first. Use to turn stable research synthesis into prose.
- 研究项目漂移状态检测Checks a shared source library and registered projects for drift: un-indexed PDFs, unmerged keywords, bibkey collisions. Use to audit a multi-project research setup.
- EMI课堂语料转写分析Segments and annotates classroom transcripts, coding language choice and discourse function with anonymised output. Use to quantify classroom talk or translanguaging.
- 人工智能伦理分析Analyzes AI ethics: algorithmic fairness metrics, bias auditing, value alignment, moral philosophy reasoning. Use to evaluate an AI system’s ethical dimensions.
- 研究计划任务执行Executes a ready research plan task by task, gating acquisition before ingest and dispatching analysis. Use after writing-research-plan to run the planned work.
- 研究项目结项清单Runs the final completion checklist: rendering, bibliography, wiki, data, reproducibility, archiving. Use when a manuscript is finalized and ready to publish.
- 形式逻辑自动推理Applies formal logic tools: Z3 SMT solver, modal logic S5, SAT encoding, Lean4 proof verification. Use to formalize or mechanically check a logical argument.
- 研究问题凝练构建Turns a broad topic or intuition into a precise, contestable, researchable question. Use at the start of a project when no defensible question yet exists.
- 历史地理信息系统分析Performs historical GIS work: georeferencing old maps, digitizing boundaries, temporal territory overlays, population interpolation. Use for historical spatial analysis.
- 历史语言学比较分析Analyzes historical linguistics in Python: sound correspondences, cognate detection, phylogenetic trees, Swadesh lists. Use when reconstructing language relationships.
- 历史社会网络分析Builds historical network analyses: correspondence networks, prosopography, temporal community detection, Gephi export. Use to map relationships among historical actors.
常见问题
人文手稿整体评审能做什么?
Performs a whole-document review of a manuscript’s question, scope, argument, citations, and structure. Use for a holistic quality check before revision。
如何安装人文手稿整体评审?
在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「人文手稿整体评审」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。
人文手稿整体评审需要哪些权限?
该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。
人文手稿整体评审经过官方审核吗?
是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。