클로드 논문 연구 프롬프트 15개 · 원문 + 번역 + 스킬 파일

한국어로 답해줘.

Read these [NUMBER OF PAPERS] papers as if you were looking for problems the authors indirectly reveal but never fully investigate. Look inside limitations, discussion sections, conflicting results, unexplained observations, assumptions, boundary conditions, and future-work sections. Generate 15 specific research questions that logically emerge from those unresolved issues. For each question, show exactly which papers created the opening and explain the reasoning in 2–3 sentences.

Focus on the Discussion, Limitations, Conclusion, and Future Research sections of these papers. Extract every concrete future-research suggestion and combine overlapping suggestions into larger research directions. For each direction give: what researchers already know, what they explicitly say is missing, which papers mention it, and 2–3 stronger research questions that could extend their suggestion. Avoid vague ideas.

Search these papers for findings that conflict with one another about [TOPIC]. Build a contradiction matrix containing: Claim A | Supporting papers | Opposing claim | Supporting papers | possible explanation for disagreement. Investigate whether differences may come from sample characteristics, geography, time period, measurements, datasets, statistical methods, definitions, or theoretical assumptions. Separate genuine contradictions from findings that only appear contradictory.

I’m uploading [NUMBER OF PAPERS] research papers about [TOPIC] in [FIELD]. Read them as one research corpus and build a structured literature review. Group the papers into 4–7 major themes, explain what each theme argues, identify the strongest evidence, show where researchers agree/disagree, and explain how thinking has evolved over time. Finish with 5 unresolved questions. Cite the specific paper/author for every important claim. Do not invent findings that are not present in the papers.

Analyze these papers on [TOPIC] specifically to find research gaps. Ignore generic gaps like “more research is needed.” Find gaps caused by contradictory findings, missing populations, weak measurements, unexplored variables, outdated datasets, methodological limitations, untested assumptions, or theories that have never been compared directly. For each gap, give: evidence from the papers → why the gap matters → a testable research question → suggested study design. Rank nothing; instead label each gap by type.

Create a knowledge map for these papers about [TOPIC]. Organize the literature into: core concepts → major theories → important variables → proposed relationships → common methods → major findings → contradictions → unresolved areas. Show how the concepts connect rather than summarizing papers individually. Then give me a text-based map using arrows such as Concept A → influences → Concept B → measured by → Method C.

Turn these papers into an evidence matrix. For every paper extract: research question, theory/framework, sample, dataset, methodology, independent variables, dependent variables, key findings, effect/direction if reported, limitations, and future research suggestions. Keep entries concise but specific. After the matrix, identify recurring methodological patterns and areas where results differ despite similar research designs.

Extract all important variables studied across these [NUMBER OF PAPERS] papers on [TOPIC]. Build a map showing which variable relationships have already been tested frequently and which plausible combinations appear rarely tested or completely untested. Then propose 10 specific relationships worth investigating in the format: X → Y, moderated/mediated by Z, with supporting reasoning based only on patterns found in the uploaded literature.

Analyze these papers chronologically from [START YEAR] to [END YEAR] and explain how research on [TOPIC] evolved. Divide the timeline into meaningful phases based on changes in theories, datasets, methodologies, assumptions, or conclusions—not arbitrary decades. For each phase explain what researchers believed, what evidence challenged those beliefs, and what new questions appeared afterward.

Examine these papers for assumptions that researchers repeatedly accept without directly testing. Look for theoretical assumptions, measurement assumptions, causal assumptions, population assumptions, data assumptions, and definitions treated as established facts. For each assumption provide: the assumption → papers relying on it → whether evidence actually supports it → how a future study could test it directly.

Using only insights from these papers about [TOPIC], create a connected research agenda containing 3–5 studies that could logically build on one another. Study 1 should establish a foundational unanswered question; later studies should extend, challenge, or generalize earlier findings. For each study provide: research question, hypothesis, theoretical contribution, variables, suggested methodology, required data, and which literature gap it addresses.

Audit the methodology used across these papers. Identify recurring weaknesses involving sample size, sampling bias, measurement quality, self-reported data, causal inference, statistical methods, replication, external validity, missing controls, dataset age, or experimental design. For every recurring weakness, cite examples from the papers and propose a stronger methodological approach a new researcher could use.

I want to study [TOPIC], but I do not want to simply repeat existing research. Based on these papers, identify 10 possible novelty angles using combinations of new context, population, theory, variable, dataset, method, mechanism, moderator, mediator, time period, or comparison. For each angle explain what existing papers already cover and exactly what the proposed study would add that is meaningfully different.

Synthesize these papers into a conceptual framework for [RESEARCH TOPIC]. Identify the most evidence-supported dependent variable, major predictors, potential mediators, moderators, control variables, and relevant theories. Show which papers support every proposed relationship. Then output the framework as: Independent Variables → Mediators → Outcome, with moderators shown separately. Clearly mark relationships that are well-established versus still uncertain.

Treat these papers as a dataset of unanswered research opportunities. Scan them across these categories: contradictions, missing populations, missing contexts, missing variables, untested mechanisms, methodological weaknesses, outdated evidence, theory conflicts, replication opportunities, unexpected findings, and unanswered author questions. Generate 20 concrete opportunities. For each one give: Opportunity | Evidence from papers | Why unresolved | Research question | Minimum study needed to investigate it. Reject any opportunity that cannot be traced back to evidence in the uploaded papers.

이 논문들로 아무도 안 한 연구 주제 찾아줘