科研 Skill
研究问题凝练构建
formulating-research-question
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.
SKILL.md
Formulating a Research Question
Purpose
Transform a broad topic or dissatisfaction into a contestable research question with an explicit object, unresolved tension, stakes, and feasible evidentiary path. This skill supports scholarly judgment; it does not manufacture novelty or decide what the researcher ought to value.
Contract
Accepts
- A topic, corpus, archive, concept, event, medium, practice, or preliminary observation.
- Notes describing what seems puzzling, inadequate, contradictory, or underexplained.
- Optional disciplinary, temporal, linguistic, geographic, or institutional constraints.
Requires
- At least one identifiable research object or candidate corpus.
- At least one unresolved tension stated without assuming its answer.
- Enough information to distinguish the topic from the research problem.
Produces
- One
ResearchQuestionrecord for the selected primary question. - Up to two subordinate questions.
- A short research-problem statement.
- A list of evidence classes required to answer the question.
- A provisional hypothesis explicitly marked
hypothesis.
May produce
- Two or three alternative question formulations with trade-offs.
- A
FAILorCONDITIONAL PASSgate report when the object, tension, or evidence path remains unclear.
Fails when
- The requested question presupposes the conclusion.
- The object or corpus cannot be identified.
- The question requires inaccessible evidence and no narrower formulation is possible.
- The request asks the agent to declare novelty without a literature check.
Guarantees
- The selected question will explicitly name its object, unresolved relation, and expected evidence path.
- Unresolved elements will remain visible rather than being silently invented.
- The output will distinguish question, hypothesis, and conclusion.
Does not guarantee
- Originality, disciplinary significance, publication value, or a successful final answer.
- That the listed evidence actually exists or will support the hypothesis.
When to use
Use this skill at the beginning of a paper, chapter, thesis, grant concept, or major revision. Use it when the researcher has many notes but no governing problem, when the current “question” is a topic label, or when the manuscript appears to know its conclusion before inquiry begins.
Do not use it merely to generate fashionable titles, inflate significance, or reverse-engineer a question around a predetermined claim. Route those cases to clarification or return FAIL.
Inputs required
Collect the following before drafting a question:
- Research object: what text, archive, discourse, practice, event, platform, institution, or concept is being examined?
- Observed feature: what is actually noticed in or about the object?
- Unresolved tension: what contradiction, absence, instability, or interpretive difficulty remains?
- Stakes: why would resolving this tension matter to a scholarly conversation?
- Evidence access: what materials can realistically be consulted?
- Constraints: discipline, language, chronology, word limit, deadline, or venue.
If one of the first three inputs is missing, do not guess it. Ask for it or produce an unresolved-items report.
Procedure
1. Separate topic from problem
Rewrite the user’s topic as a neutral noun phrase. Then state the unresolved problem as a full sentence. A topic such as “platform aesthetics” is not yet a problem. A problem appears only when an observed feature and an unresolved tension are connected.
Use the grammar:
ResearchProblem = Observation + Unresolved Tension + Stakes
2. Identify the unit of analysis
Specify whether the inquiry concerns a text, motif, interface, archive, historical episode, concept, discourse, reception pattern, institutional practice, or relation among these. Avoid shifting units mid-question.
3. Generate question families
Draft at least three candidates when enough input exists:
- Interpretive: How should a difficult feature or relation be understood?
- Explanatory-genealogical: How did a concept, form, or discourse acquire its present configuration?
- Critical-comparative: What becomes visible when two positions, corpora, or media forms are placed in a bounded comparison?
Do not force all three if the object clearly supports only one.
4. Test answerability
For each candidate, ask:
- Can the question be answered with identifiable materials?
- Is the answer non-obvious and contestable?
- Could evidence require revision of the initial hypothesis?
- Is the scale appropriate for the venue and deadline?
- Does the wording avoid evaluative slogans such as “bad,” “revolutionary,” or “destroyed” unless these are objects of analysis?
5. Detect conclusion smuggling
Mark and revise formulations that contain their preferred answer. For example, “How does platform capitalism destroy authentic experience?” already treats destruction and authenticity as settled. Reframe the relation as an open problem.
6. Select and register
Select one primary question. Record:
- object
- unresolved relation
- scope cues
- evidence classes
- stakes
- verification state
Assign verification_state: researcher-supplied unless relevant claims have been independently checked.
7. Draft a provisional hypothesis
Only after the question is stable, state a provisional answer. Label it hypothesis. Add at least one condition under which the hypothesis would need revision.
8. Revisit the question after exploratory source work
Treat question formation as iterative rather than final. After an initial source, corpus, archive, or concept exploration, compare the provisional question with what the materials actually make visible. Record whether the object, tension, scale, vocabulary, or evidence path has changed. When exploratory materials expose a different problem, route back to this skill instead of forcing the original question to survive.
A revised question MUST preserve a change note:
Original question
→ exploratory finding
→ reason for revision
→ revised question
Do not describe this return as failure. In humanities inquiry, revision after reading is often evidence that the question has become more precise.
9. Issue the gate report
Return PASS only when object, tension, stakes, and evidence path are explicit. Return CONDITIONAL PASS when the question is usable but one non-blocking scope or terminology decision remains. Return FAIL when the question is circular, inaccessible, or too indeterminate to hand off.
Stop signals
Stop and preserve uncertainty when:
- the user asks for a “groundbreaking” question without a corpus or literature context;
- the question embeds an unverified historical claim;
- the key concept has no stable working definition;
- the intended evidence is unavailable, private, untranslated, or outside the project’s capacity;
- the requested scale would require multiple independent studies;
- the researcher rejects every formulation that does not already prove the preferred conclusion.
When a stop signal occurs, produce Unresolved items and Researcher decision required. Do not repair the gap by inventing a source, consensus, or historical premise.
Completion criteria
The skill is complete only when:
- the primary question contains an identifiable object and analytical relation;
- the unresolved tension is not merely a topic label;
- the answer could plausibly be contested;
- the evidence classes are named;
- the scale is feasible under stated constraints;
- the provisional hypothesis is marked as revisable;
- the gate report records blocking and non-blocking issues;
- the next skill can accept the produced object without reconstructing missing context.
Anti-fabrication rules
- Do not declare a question original without a completed literature dialogue.
- Do not invent gaps in scholarship.
- Do not convert the researcher’s intuition into a verified disciplinary fact.
- Do not invent quotations, publication facts, page numbers, DOI values, archival identifiers, or source contents.
- Do not imply that a source exists merely because a plausible title can be generated.
- Preserve disagreement and uncertainty.
- Use
unknownwhen the relevant fact has not been checked.
Output format
## Research problem
Observation:
Unresolved tension:
Stakes:
## Candidate questions
### Candidate A
Question:
Type:
Strength:
Risk:
Evidence required:
## Selected primary question
Question:
Object:
Analytical relation:
Verification state:
## Sub-questions
1.
2.
## Provisional hypothesis
Status: hypothesis
Statement:
Revision condition:
## Unresolved items
## Gate report
Gate: Research-question gate
Status: PASS | CONDITIONAL PASS | FAIL
Evidence checked:
Blocking issues:
Non-blocking risks:
Researcher decisions required:
Required next action:
Next skill: scoping-argument-boundary
Good invocation
Use this skill to turn my topic “platform aesthetics” and my notes on artificial nature into three candidate research questions. My corpus is a set of museum platform interfaces from 2023–2025. Explain the trade-offs and select one question only if the evidence path is feasible.
This invocation supplies an object, a provisional tension, a bounded corpus, and a request for transparent selection.
Bad invocation
Give me a groundbreaking research question that proves platforms have destroyed nature. Make it sound publishable and say nobody has studied it.
This request presupposes a conclusion, asks for unsupported novelty, and provides no bounded object. The correct outcome is clarification or FAIL, not an impressive-sounding question.
Next skills
scoping-argument-boundaryto delimit corpus, period, comparison, and excluded claims.mapping-concept-lineagewhen key terms require genealogy and differentiation.conducting-literature-dialoguebefore any originality claim is accepted.
Friction triggers
Create premise friction when a preferred answer is embedded in the question, an unverified premise controls the inquiry, or new material undermines the question's object or tension.
Friction checks
Compare the premise, evidence path, and revision conditions with exploratory findings. Record CLAIM_OVERREACH, INFERENCE_PROMOTION, or EVIDENCE_CONFLICT only when the issue materially affects answerability.
Judgment boundary
The agent may present question alternatives and consequences; the researcher selects the governing question and stakes. Preserve recommendations separately from the authorized decision.
Evidence ledger updates
Register evidence classes and any source tied to a premise. Verification state does not determine whether that source supports the premise.
Interpretation history impact
When a new reading changes the question, link the accepted or proposed interpretation version rather than replacing the earlier rationale.
Possible epistemic return
This skill is a primary return target when new sources or interpretations materially alter an earlier question. Preserve Q1, create Q2 with supersedes, and identify affected scope, argument, and submission gates.
Productive refusal
If originality, first use, or a premise cannot be established, report the checked range, missing search, bounded wording currently permitted, and next verification action.
Gate impact
Premise friction that makes the question circular or unsupported forces FAIL. A materially revised question marks dependent gates INVALIDATED or REQUIRES_RECHECK without changing their historical status.
Limitations
A well-formed question can still be unimportant, derivative, ethically unsuitable, or impossible to answer with the eventual evidence. Question quality must be reassessed after literature review, close reading, and argument stress testing.
示例提示词
我有一个宽泛的研究方向,帮我把它凝练成一个可辩驳的具体研究问题。
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.
- 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.
- 历史文献OCR识别Transcribes printed and handwritten historical documents with Tesseract/Kraken OCR, confidence filtering, spell correction. Use for digitizing scanned historical text.
常见问题
研究问题凝练构建能做什么?
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。
如何安装研究问题凝练构建?
在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「研究问题凝练构建」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。
研究问题凝练构建需要哪些权限?
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。
研究问题凝练构建经过官方审核吗?
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。