{"id":"f5455555-082b-4d5d-808c-a06fd53ae5d0","shortId":"NJttS5","kind":"skill","title":"prompt-builder","tagline":"Interactive structured prompt generator for Claude based on the user's task description. Asks 7 questions and assembles a ready-to-use prompt from a universal template. EN triggers: 'create a prompt', 'write a prompt', 'prompt builder'. RU triggers: 'создай промпт', 'напиши промп","description":"# Prompt Builder\n\nAn interactive generator that builds structured prompts for Claude based on the user's task description. The skill asks 7 questions, interprets loose or vague answers, clarifies details, and assembles a ready-to-use prompt from a universal template.\n\nSuitable for anyone who needs to create a reusable, well-structured prompt for Claude — for code review, content writing, data analysis, or any other repeatable task.\n\n## Language Detection\n\nDetect the language of the user's first message:\n- If the message is in **Russian** — conduct the entire interview and output the final prompt in Russian.\n- If the message is in **English** — conduct the entire interview and output the final prompt in English.\n- If the language is ambiguous, default to English.\n\nAll question texts, labels, and output section headers must match the detected language.\n\n## Input\n\nA natural-language description from the user of what they want the prompt to accomplish. Can be brief (\"I need a prompt for code review\") or detailed.\n\n## Output\n\nA structured prompt document with sections: Role, Context, Task, Input data, Output requirements, Constraints, and optionally Examples. Displayed in chat and optionally saved as a `.md` file.\n\n## Instructions\n\n### Step 1: Interactive interview\n\nAsk questions in the order below. Wait for the user's answer after each question before proceeding.\n\n**Rules:**\n- User answers may be unstructured or vague — interpret the intent\n- If an answer is unclear or incomplete, ask a clarifying question\n- Do not move to the next item until you have enough information\n- Collect answers into variables for final generation\n\n---\n\n#### Question 1: Claude's role\n\n**EN:** \"What role should Claude take on in this task? (E.g.: analyst, writer, tester, coder, etc.) Describe what it will do in general terms.\"\n\n**RU:** \"Какую роль должен взять на себя Claude в этой задаче? (Например: аналитик, писатель, тестировщик, кодер и т.д.) Опиши, что он будет делать в общих чертах.\"\n\n**Action:** Interpret the answer, extract the role and core function. If the answer is too broad, ask: \"If I understood correctly, Claude should act as [your interpretation]? Confirm or clarify?\"\n\n**Variable:** `ROLE`\n\n---\n\n#### Question 2: Context\n\n**EN:** \"What context does Claude need to understand the task? (Background: why this matters, conditions, prerequisites?)\"\n\n**RU:** \"Какой контекст нужен Claude для понимания задачи? (Фоновая информация: почему это важно, в каких условиях, какие предпосылки?)\"\n\n**Action:** Interpret as background and prerequisites. If the user says \"it's clear from the task\", ask: \"Are there specific conditions, constraints, or history that affect the task?\"\n\n**Variable:** `CONTEXT`\n\n---\n\n#### Question 3: Main task\n\n**EN:** \"Specifically, what should Claude produce? (Describe the exact result you need)\"\n\n**RU:** \"Конкретно, что должен выполнить Claude? (Опиши точный результат, который нужен)\"\n\n**Action:** Interpret as the primary goal. If the answer is vague, reframe: \"So you need [your interpretation]?\"\n\n**Variable:** `TASK`\n\n---\n\n#### Question 4: Input data\n\n**EN:** \"What data or information will be passed to Claude? (Text, table, list, description, nothing?)\"\n\n**RU:** \"Какие данные/информация будут подаваться на вход Claude? (Текст, таблица, список, описание, ничего?)\"\n\n**Action:** Interpret the format and type of input. If \"it depends\", ask: \"Give an example of a typical input.\"\n\n**Variable:** `INPUT`\n\n---\n\n#### Question 5: Output requirements\n\n**EN:** \"What should the result look like? (Format: text, list, table, code, JSON, etc.) (Length: brief, detailed, specific number of items?) (Style: technical, plain language, with examples?)\"\n\n**RU:** \"В каком виде должен быть результат? (Формат: текст, список, таблица, код, JSON и т.д.) (Объем: краткий, развёрнутый, конкретное количество пунктов?) (Стиль: технический, простой язык, с примерами?)\"\n\n**Action:** Interpret all three aspects. If only one is answered, ask about the others: \"And the format? Length? Style?\"\n\n**Variable:** `OUTPUT`\n\n---\n\n#### Question 6: Constraints and tone\n\n**EN:** \"Are there any constraints or special requirements? (What to avoid, tone, taboos, formatting restrictions?)\"\n\n**RU:** \"Есть ли ограничения или специальные требования? (Что нельзя делать, особый тон, табу, форматирование?)\"\n\n**Action:** Interpret as constraints and stylistic requirements. If \"no constraints\", ask: \"Can Claude be creative? Is there a preferred style (formal/informal)?\"\n\n**Variable:** `CONSTRAINTS`\n\n---\n\n#### Question 7: Examples\n\n**EN:** \"Do you need input/output examples for clarity? (If yes, provide one: what goes in, what's expected out)\"\n\n**RU:** \"Нужны ли примеры input/output для ясности? (Если да, приведи пример: что на входе, что ожидается на выходе)\"\n\n**Action:** If \"yes\" — ask for an example. If \"no\" — skip and proceed to generation.\n\n**Variable:** `EXAMPLES` (optional)\n\n---\n\n### Step 2: Generate the prompt\n\nAfter collecting all answers, assemble the final prompt using the appropriate template.\n\n**EN template:**\n```\n## Role\n[ROLE]\n\n## Context\n[CONTEXT]\n\n## Task\n[TASK]\n\n## Input data\n[INPUT]\n\n## Output requirements\n[OUTPUT]\n\n## Constraints\n[CONSTRAINTS]\n\n[IF EXAMPLES COLLECTED:]\n## Examples\n[EXAMPLES]\n```\n\n**RU template:**\n```\n## Роль\n[ROLE]\n\n## Контекст\n[CONTEXT]\n\n## Задача\n[TASK]\n\n## Входные данные\n[INPUT]\n\n## Требования к выводу\n[OUTPUT]\n\n## Ограничения\n[CONSTRAINTS]\n\n[ЕСЛИ EXAMPLES СОБРАНЫ:]\n## Примеры\n[EXAMPLES]\n```\n\n---\n\n### Step 3: Output and save\n\n1. Display the final prompt in chat — clearly and structured\n2. After the prompt, ask the user: \"The prompt is ready. Do you want to save it as a `.md` file?\" (in the detected language)\n3. If \"yes\":\n   - Suggest a filename (e.g.: `prompt-[short-description].md`)\n   - Save the file and make it available for download or copying\n4. If \"no\":\n   - Ask: \"Do you need any edits to the prompt?\"\n   - If edits — return to the relevant step, update the variable, regenerate\n\n## Output Format\n\nThe generated prompt is a Markdown document with H2-level section headers. Each section contains only the user-provided content for that field, lightly formatted for readability. The document is self-contained and ready to paste directly into a new Claude conversation.\n\n## Negative Cases\n\n- **User provides a full, detailed description upfront** — still ask all 7 questions to confirm each field; prefill your interpretation and ask the user to confirm rather than skipping\n- **User answers in a different language mid-interview** — switch to that language for remaining questions and the final output\n- **User wants to skip a question** — accept \"skip\" or \"not applicable\" and use a reasonable default (e.g., \"No specific constraints\") rather than blocking\n- **User asks to edit the generated prompt** — return to the relevant question(s), update variables, and regenerate the full prompt\n- **User provides contradictory answers** — ask for clarification rather than guessing which answer to use","tags":["prompt","builder","claude","skills","kit","kirkruglov","agent-skills","agentic-skills","ai-agents","ai-skills","awesome-list","claude-ai"],"capabilities":["skill","source-kirkruglov","skill-prompt-builder","topic-agent-skills","topic-agentic-skills","topic-ai-agents","topic-ai-skills","topic-awesome-list","topic-claude","topic-claude-ai","topic-claude-ai-skills","topic-claude-code","topic-claude-cowork","topic-claude-memory","topic-claude-skills"],"categories":["claude-skills-kit"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/KirKruglov/claude-skills-kit/prompt-builder","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add KirKruglov/claude-skills-kit","source_repo":"https://github.com/KirKruglov/claude-skills-kit","install_from":"skills.sh"}},"qualityScore":"0.453","qualityRationale":"deterministic score 0.45 from registry signals: · indexed on github topic:agent-skills · 7 github stars · SKILL.md body (6,983 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-05-18T19:13:38.375Z","embedding":null,"createdAt":"2026-05-18T13:21:15.984Z","updatedAt":"2026-05-18T19:13:38.375Z","lastSeenAt":"2026-05-18T19:13:38.375Z","tsv":"'1':242,304,817 '2':392,753,827 '3':459,813,852 '4':505,875 '5':559 '6':639 '7':18,69,696,957 'accept':1001 'accomplish':199 'act':382 'action':359,428,485,537,617,672,735 'affect':453 'ambigu':166 'analysi':111 'analyst':319 'answer':75,256,264,275,297,362,371,493,626,760,976,1041,1049 'anyon':92 'applic':1005 'appropri':767 'ask':17,68,245,280,375,444,548,627,682,738,831,878,955,967,1019,1042 'aspect':621 'assembl':21,79,761 'avail':870 'avoid':653 'background':404,431 'base':10,59 'block':1017 'brief':202,577 'broad':374 'build':54 'builder':3,41,49 'case':946 'chat':232,823 'clarif':1044 'clarifi':76,282,388 'clariti':705 'claud':9,58,104,305,312,339,380,398,414,466,479,517,531,684,943 'clear':440,824 'code':106,208,573 'coder':322 'collect':296,758,787 'condit':408,448 'conduct':134,151 'confirm':386,960,971 'constraint':226,449,640,647,675,681,694,783,784,806,1014 'contain':915,934 'content':108,921 'context':220,393,396,457,773,774,795 'contradictori':1040 'convers':944 'copi':874 'core':367 'correct':379 'creat':34,96 'creativ':686 'data':110,223,507,510,778 'default':167,1010 'depend':547 'describ':324,468 'descript':16,65,188,521,862,952 'detail':77,211,578,951 'detect':118,119,181,850 'differ':979 'direct':939 'display':230,818 'document':216,906,930 'download':872 'e.g':318,858,1011 'edit':883,888,1021 'en':32,308,394,462,508,562,643,698,769 'english':150,161,169 'enough':294 'entir':136,153 'etc':323,575 'exact':470 'exampl':229,551,588,697,703,741,750,786,788,789,808,811 'expect':715 'extract':363 'field':924,962 'file':239,847,866 'filenam':857 'final':141,158,301,763,820,993 'first':126 'formal/informal':692 'format':540,569,633,656,899,926 'full':950,1036 'function':368 'general':330 'generat':7,52,302,748,754,901,1023 'give':549 'goal':490 'goe':711 'guess':1047 'h2':909 'h2-level':908 'header':177,912 'histori':451 'incomplet':279 'inform':295,512 'input':183,222,506,544,555,557,777,779,800 'input/output':702,721 'instruct':240 'intent':272 'interact':4,51,243 'interpret':71,270,360,385,429,486,501,538,618,673,965 'interview':137,154,244,983 'item':290,582 'json':574,601 'label':173 'languag':117,121,164,182,187,586,851,980,987 'length':576,634 'level':910 'light':925 'like':568 'list':520,571 'look':567 'loos':72 'main':460 'make':868 'markdown':905 'match':179 'matter':407 'may':265 'md':238,846,863 'messag':127,130,147 'mid':982 'mid-interview':981 'move':286 'must':178 'natur':186 'natural-languag':185 'need':94,204,399,473,499,701,881 'negat':945 'new':942 'next':289 'noth':522 'number':580 'one':624,709 'option':228,234,751 'order':249 'other':630 'output':139,156,175,212,224,560,637,780,782,804,814,898,994 'pass':515 'past':938 'plain':585 'prefer':690 'prefil':963 'prerequisit':409,433 'primari':489 'proceed':261,746 'produc':467 'prompt':2,6,27,36,39,40,48,56,85,102,142,159,197,206,215,756,764,821,830,835,859,886,902,1024,1037 'prompt-build':1 'provid':708,920,948,1039 'question':19,70,171,246,259,283,303,391,458,504,558,638,695,958,990,1000,1029 'rather':972,1015,1045 'readabl':928 'readi':24,82,837,936 'ready-to-us':23,81 'reason':1009 'refram':496 'regener':897,1034 'relev':892,1028 'remain':989 'repeat':115 'requir':225,561,650,678,781 'restrict':657 'result':471,566 'return':889,1025 'reusabl':98 'review':107,209 'role':219,307,310,365,390,771,772,793 'ru':42,332,410,474,523,589,658,717,790 'rule':262 'russian':133,144 'save':235,816,842,864 'say':437 'section':176,218,911,914 'self':933 'self-contain':932 'short':861 'short-descript':860 'skill':67 'skill-prompt-builder' 'skip':744,974,998,1002 'source-kirkruglov' 'special':649 'specif':447,463,579,1013 'step':241,752,812,893 'still':954 'structur':5,55,101,214,826 'style':583,635,691 'stylist':677 'suggest':855 'suitabl':90 'switch':984 'tabl':519,572 'taboo':655 'take':313 'task':15,64,116,221,317,403,443,455,461,503,775,776,797 'technic':584 'templat':31,89,768,770,791 'term':331 'tester':321 'text':172,518,570 'three':620 'tone':642,654 'topic-agent-skills' 'topic-agentic-skills' 'topic-ai-agents' 'topic-ai-skills' 'topic-awesome-list' 'topic-claude' 'topic-claude-ai' 'topic-claude-ai-skills' 'topic-claude-code' 'topic-claude-cowork' 'topic-claude-memory' 'topic-claude-skills' 'trigger':33,43 'type':542 'typic':554 'unclear':277 'understand':401 'understood':378 'univers':30,88 'unstructur':267 'updat':894,1031 'upfront':953 'use':26,84,765,1007,1051 'user':13,62,124,191,254,263,436,833,919,947,969,975,995,1018,1038 'user-provid':918 'vagu':74,269,495 'variabl':299,389,456,502,556,636,693,749,896,1032 'wait':251 'want':195,840,996 'well':100 'well-structur':99 'write':37,109 'writer':320 'yes':707,737,854 'аналитик':344 'будет':354 'будут':527 'быть':594 'в':340,356,423,590 'важно':422 'взять':336 'виде':592 'вход':530 'входе':730 'входные':798 'выводу':803 'выполнить':478 'выходе':734 'д':350,604 'да':725 'данные':525,799 'делать':355,667 'для':415,722 'должен':335,477,593 'если':724,807 'есть':659 'задача':796 'задаче':342 'задачи':417 'и':348,602 'или':662 'информация':419,526 'к':802 'какие':426,524 'каких':424 'какой':411 'каком':591 'какую':333 'код':600 'кодер':347 'количество':609 'конкретно':475 'конкретное':608 'контекст':412,794 'который':483 'краткий':606 'ли':660,719 'на':337,529,729,733 'напиши':46 'например':343 'нельзя':666 'ничего':536 'нужен':413,484 'нужны':718 'общих':357 'объем':605 'ограничения':661,805 'ожидается':732 'он':353 'описание':535 'опиши':351,480 'особый':668 'писатель':345 'подаваться':528 'понимания':416 'почему':420 'предпосылки':427 'приведи':726 'пример':727 'примерами':616 'примеры':720,810 'промп':47 'промпт':45 'простой':613 'пунктов':610 'развёрнутый':607 'результат':482,595 'роль':334,792 'с':615 'себя':338 'собраны':809 'создай':44 'специальные':663 'список':534,598 'стиль':611 'т':349,603 'таблица':533,599 'табу':670 'текст':532,597 'тестировщик':346 'технический':612 'тон':669 'точный':481 'требования':664,801 'условиях':425 'фоновая':418 'формат':596 'форматирование':671 'чертах':358 'что':352,476,665,728,731 'это':421 'этой':341 'язык':614 'ясности':723","prices":[{"id":"ede721ce-e783-49b3-81f1-14d083964650","listingId":"f5455555-082b-4d5d-808c-a06fd53ae5d0","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"KirKruglov","category":"claude-skills-kit","install_from":"skills.sh"},"createdAt":"2026-05-18T13:21:15.984Z"}],"sources":[{"listingId":"f5455555-082b-4d5d-808c-a06fd53ae5d0","source":"github","sourceId":"KirKruglov/claude-skills-kit/prompt-builder","sourceUrl":"https://github.com/KirKruglov/claude-skills-kit/tree/main/skills/prompt-builder","isPrimary":false,"firstSeenAt":"2026-05-18T13:21:15.984Z","lastSeenAt":"2026-05-18T19:13:38.375Z"}],"details":{"listingId":"f5455555-082b-4d5d-808c-a06fd53ae5d0","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"KirKruglov","slug":"prompt-builder","github":{"repo":"KirKruglov/claude-skills-kit","stars":7,"topics":["agent-skills","agentic-skills","ai","ai-agents","ai-skills","awesome-list","claude","claude-ai","claude-ai-skills","claude-code","claude-cowork","claude-memory","claude-skills","memory-management","productivity","productivity-tools","project-management"],"license":"mit","html_url":"https://github.com/KirKruglov/claude-skills-kit","pushed_at":"2026-05-18T04:27:46Z","description":"40+ curated agent skills for Claude Cowork and Claude.ai — ready-to-use tools for non-technical users: project management, productivity, and AI workflow automation","skill_md_sha":"a30173cf51f78e71c96efcd63528b3df7c01dc9e","skill_md_path":"skills/prompt-builder/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/KirKruglov/claude-skills-kit/tree/main/skills/prompt-builder"},"layout":"multi","source":"github","category":"claude-skills-kit","frontmatter":{"name":"prompt-builder","description":"Interactive structured prompt generator for Claude based on the user's task description. Asks 7 questions and assembles a ready-to-use prompt from a universal template. EN triggers: 'create a prompt', 'write a prompt', 'prompt builder'. RU triggers: 'создай промпт', 'напиши промпт', 'prompt builder'."},"skills_sh_url":"https://skills.sh/KirKruglov/claude-skills-kit/prompt-builder"},"updatedAt":"2026-05-18T19:13:38.375Z"}}