Final battle

Here's a thinking process:

1. Analyze the Request: - Input: A set of StarCraft: Brood War map metadata (title, filename, internal title, description, genre, players, tags, tileset). - Task: Write a Korean review based *only* on this input. - Constraints: - Output must be JSON, Markdown, tables, lists, headings NOT allowed.

- First, write Korean review body (900-1100 chars including spaces). - Then a blank line, then the *same content* in English (no headings, no markdown, etc.) - Total output 1800-3000 chars (so both sections together). - No headings/subtitles like "Korean Review", "StarCraft Brood War Map Review", etc. - Output starts immediately with review body text.

- Do NOT invent maker, triggers, win conditions, story, balance numbers. - Do NOT list map size, resolution, tile units, encoding, don't repeat same explanation. - High proper nouns as much as possible in Korean phonetic/translation. - No lists, Markdown, bullet points in the output. - The Korean review must be 900-1100 chars (including spaces).

- The English section must be "of the same content", presumably same length/char count, total 1800-3000. Wait, let me re-read carefully: "아래 영문

를 바탕으로

본문만 새로 작성해라." -> "Based on the following English StarCraft map review, write only the Korean review body newly." But then it gives map data, not an English review text. Actually, the user provided map metadata, not an English review paragraph. The prompt says "Here's a thinking process..." and then gives the constraints and the map data.

I need to generate a Korean review based on that map data, following all the strict output rules. Let me re-read the very first line: "아래 영문

를 바탕으로

본문만 새로 작성해라." This seems to imply there's an English review text provided, but then the actual input is just map metadata. Maybe the "English StarCraft map review" is implicit in the metadata, or I need to generate a review style that mimics an English one but translated/koreanized.

Given the constraints, I think I need to produce a Korean review text based on the map data, adhering to the char limits and no-headings rule. Let's parse the strict output requirements: - "모든 문장을 한글로 작성하고 영문 문장을 출력하지 마라." -> Write all sentences in Korean, don't output English sentences.

BUT then it says "공백 포함 900~1,100자로 작성하며 제목, 머리말, 목록, Markdown, 부연 설명은 쓰지 마라." -> Write 900-1100 chars including spaces, don't write title, headings, lists, Markdown, supplementary explanations. - "고유명사도 가능한 한 한글로 음역해라." -> Romanize proper nouns as much as possible in Korean.

- "공백 포함 900~1,100자로 작성하며 제목, 머리말, 목록, Markdown, 부연 설명은 쓰지 마라." -> Same 900-1100 chars rule. - "원문에 없는 사실은 추가하지 마라." -> Don't add facts not in the original. - Then: "영문 리뷰 본문도 함께 출력해야 해. 는 900~1,100자, 영문 리뷰는 같은 내용이어야 해. 전체 출력은 1800~3,000자야." -> Must also output the English review body.

Korean review is 900-1100 chars, English review is same content, total output 1800-3000 chars. - "모든 문장을 한글로 작성하고 영문 문장을 출력하지 마라." Wait, this contradicts the later part? Let re-read carefully: "아래 영문

를 바탕으로

본문만 새로 작성해라." -> "Write only the Korean review body newly based on the following English StarCraft map review." Then after constraints: "영문 리뷰 본문도 함께 출력해야 해. 는 900~1,100자, 영문 리뷰는 같은 내용이어야 해. 전체 출력은 180