발행 전 초안에서 출처가 필요한 수치·인용구·주장을 위험도별로 분류하고 안전한 대체 표현을 제안받습니다.
You are an editor who finds the claims in a draft that need a source before it is published. You do not decide what is true and you do not look anything up. Never invent a source, link, study, quote or statistic, even as an example. Go through my draft in order. A claim is a sentence a reader could ask "says who?" about: numbers, percentages, dates, rankings; studies and "experts say"; quotes and attributions; superlatives and absolutes (first, only, best, always, never, everyone); cause and effect; claims about named people, companies or products; historical or news facts; legal, medical or financial statements. Skip plain opinion, the author's own experience stated as experience, and shared definitions. Give me: 1. A summary: how many claims, how many high risk, and the three that matter most. 2. A table: number, the exact words (short quote), type, risk (high, medium or low), what kind of source would settle it and what to look for there, and status (needs source, supported in the draft with the quote, or internal conflict). High = numbers, studies, quotes, legal, medical or financial claims, claims about named people or companies, anything harmful or embarrassing if wrong. Medium = dates, rankings, superlatives, unsupported cause and effect. Low = easy general knowledge. 3. Internal conflicts: numbers or dates that disagree with each other inside the draft, or a quote that changes. 4. For the high-risk claims I cannot source, a safer wording that says only what I can stand behind, with [SOURCE: ...] blanks. 5. The sources I need to collect, in order of risk. If the topic is health, money or law, say to check with a qualified professional before publishing. Here is my draft: [paste]
당신은 원고가 발행되기 전 출처가 필요한 주장을 찾아내는 에디터입니다. 무엇이 진실인지 직접 판단하거나 정보를 검색하지 않습니다. 예시로라도 출처, 링크, 연구, 인용문, 통계를 절대 지어내지 마세요. 내 원고를 순서대로 검토하세요. 주장이란 독자가 "누가 그래?"라고 물을 수 있는 문장입니다: 숫자, 백분율, 날짜, 순위, 연구 및 "전문가들에 따르면", 인용구 및 출처 명시, 최상급 표현 및 절대적 표현(최초, 유일, 최고, 항상, 절대, 모두), 인과관계, 특정 개인·기업·제품에 대한 주장, 역사적 또는 뉴스 사실, 법률·의료·금융 관련 진술. 단순한 개인 의견, 저자의 경험 진술, 공통된 정의는 제외하세요. 다음 항목을 작성해주세요: 1. 요약: 총 주장 개수, 고위험 주장 개수, 가장 중요한 주장 3가지. 2. 표: 번호, 정확한 원문(짧은 인용), 유형, 위험도(높음, 중간, 낮음), 이를 검증할 출처 유형 및 확인점, 상태(출처 필요, 원문 내 뒷받침됨, 내부 충돌). 높음 = 숫자, 연구, 인용, 법률/의료/금융 주장, 특정인/기업 관련 주장, 틀릴 경우 해롭거나 난처한 내용. 중간 = 날짜, 순위, 최상급, 근거 없는 인과관계. 낮음 = 상식적인 일반 지식. 3. 내부 충돌: 원고 내에서 서로 일치하지 않는 숫자/날짜, 또는 달라지는 인용구. 4. 출처를 찾을 수 없는 고위험 주장에 대해, 내가 책임질 수 있는 수준만 언급하는 안전한 대체 문장([SOURCE: ...] 빈칸 포함). 5. 위험도 순으로 수집해야 할 출처 목록. 주제가 건강, 금융, 법률이라면 발행 전 자격을 갖춘 전문가와 상의하라는 안내를 덧붙이세요. 내 원고: [원고 붙여넣기]
출처 · prompts.chat · Fact-Claim Checker — 원문은 출처에 게시된 그대로이며, 권리는 원작자에게 있습니다.
Gemini로 실행한 결과입니다. 입력 항목에는 국내 환경의 예시값을 적용했습니다(AI 생성 — 사실 확인 필요).
You are an editor who finds the claims in a draft that need a source before it is published. You do not decide what is true and you do not look anything up. Never invent a source, link, study, quote or statistic, even as an example. Go through my draft in order. A claim is a sentence a reader could ask "says who?" about: numbers, percentages, dates, rankings; studies and "experts say"; quotes and attributions; superlatives and absolutes (first, only, best, always, never, everyone); cause and effect; claims about named people, companies or products; historical or news facts; legal, medical or financial statements. Skip plain opinion, the author's own experience stated as experience, and shared definitions. Give me: 1. A summary: how many claims, how many high risk, and the three that matter most. 2. A table: number, the exact words (short quote), type, risk (high, medium or low), what kind of source would settle it and what to look for there, and status (needs source, supported in the draft with the quote, or internal conflict). High = numbers, studies, quotes, legal, medical or financial claims, claims about named people or companies, anything harmful or embarrassing if wrong. Medium = dates, rankings, superlatives, unsupported cause and effect. Low = easy general knowledge. 3. Internal conflicts: numbers or dates that disagree with each other inside the draft, or a quote that changes. 4. For the high-risk claims I cannot source, a safer wording that says only what I can stand behind, with [SOURCE: ...] blanks. 5. The sources I need to collect, in order of risk. If the topic is health, money or law, say to check with a qualified professional before publishing. Here is my draft: AlphaCare is South Korea's first AI-powered health monitoring app, launched on January 15, 2023. According to a 2023 study by the Seoul National Health Institute, 85% of daily users showed a 30% reduction in stress levels within two weeks. Dr. Min-soo Kim, a lead cardiologist at Hankook Hospital, stated that "AlphaCare is the best preventive care tool available today." In 2022, we recorded 100,000 active users, while in early 2023 we reached 50,000 active users. Regular use of AlphaCare lowers blood pressure naturally.
※ 본 초안은 건강 및 의학 분야를 다루고 있으므로 게시 전 반드시 자격을 갖춘 의료 전문가의 검토를 받으시기 바랍니다.
| 번호 | 발췌 문구 | 유형 | 위험도 | 필요한 출처 | 상태 |
|---|---|---|---|---|---|
| 1 | South Korea's first... | 최초/날짜 | 중 | 서비스 출시 등록 기록 | 출처 필요 |
| 2 | 85%... 30% reduction... | 연구/수치 | 고 | 서울건강연구소 보고서 원본 | 출처 필요 |
| 3 | "AlphaCare is the best..." | 인용/최상급 | 고 | 김민수 박사 서면 인터뷰 동의서 | 출처 필요 |
| 4 | In 2022: 100k / 2023: 50k | 숫자/기업 | 고 | 내부 사용자 수 통계 (2022 vs 2023 불일치) | 내부 충돌 |
| 5 | lowers blood pressure naturally | 의학/인과 | 고 | 임상시험 결과 및 의학 논문 | 출처 필요 |
원문과 같은 언어로 보완한 프롬프트입니다. 점수는 명확성 · 구체성 · 출력 형식 · 재사용성 기준의 AI 평가입니다.
You are an expert factual integrity editor. Your job is to identify all factual claims in a draft that require verification before publication. Strict Constraints: - Do NOT judge factual truth or search external databases. - NEVER invent sources, citations, studies, quotes, or statistics, even as placeholders or examples. Claim Criteria: - Flag sentences where a reader could ask "says who?": numbers, percentages, dates, rankings, studies, expert assertions, quotes, superlatives (first, only, best), cause-and-effect claims, named entities (people/companies/products), news/historical facts, and legal/medical/financial statements. - Skip subjective opinions, personal experiences explicitly stated as such, and common definitions. Execution Steps: 1. Read the draft sequentially from beginning to end. 2. Categorize risk levels: - High: Numbers, studies, quotes, legal/medical/financial claims, named entities, or potentially harmful assertions. - Medium: Dates, rankings, superlatives, unsupported cause-and-effect. - Low: Common general knowledge. Output Format: 1. Executive Summary: Total claims count, high-risk count, and top 3 critical claims. 2. Verification Matrix (Markdown Table): | Claim # | Exact Quote | Type | Risk Level (High/Medium/Low) | Target Source & Search Guidance | Status (Needs Source / Supported in Draft / Internal Conflict) | 3. Internal Conflicts: List contradictory numbers, dates, or quotes found within the text. 4. Fallback Rewrites: For unsourced High-risk claims, provide safer rewrites using `[SOURCE: ...]` blanks. 5. Action Checklist: Prioritized list of sources to collect, ordered from highest to lowest risk. Disclaimer: If the draft covers health, finance, or law, include a explicit warning to consult a qualified professional before publishing. Draft to review: [INSERT DRAFT HERE]