Untitled Scenario.

Here's a thinking process:

1. Analyze the Request: - Role: StarCraft: Broodwar map reviewer. - Source Data: Provided map data (title, filename, internal title, internal description, genre, players, teams, playtime, difficulty, version, tags, tileset). - Output Rules: - No JSON, Markdown code blocks, tables, lists, headings. - First: Korean review, 900-1100 characters (with spaces).

- Then, a blank line, then the same content in English (but the prompt says "Korean translation本本" - wait, let me re-read carefully). - Actually, the prompt says: "다음 를 자연스러운

로 번역해라. 제목, 머리말, 목록, Markdown, 부연 설명 없이 한국어 번역 본문만 출력해라. 원문에 없는 맵 정보나 평가를 추가하지 마라." - Then there's a lot of confusing formatting about "11:15 AM length: 900~1100 chars... then a blank line, then the same content no subtitles/headings... Total output: 1800~3000 chars (so Korean + English + blank line)." - Wait, let me re-read the entire prompt carefully.

It's a mix of English and Korean constraints. Key parts: - "다음 를 자연스러운

로 번역해라." -> Translate the following SC map review into a natural Korean review. - "제목, 머리말, 목록, Markdown, 부연 설명 없이 한국어 번역 본문만 출력해라." -> Output only the Korean translation body, without title, headings, lists, Markdown, extra explanations. - "원문에 없는 맵 정보나 평가를 추가하지 마라." -> Don't add map info or evaluations not in the original.

- Then there's: "Here's a thinking process:" and a long AI instruction block. - The instruction block says: "- 11:15 AM length: 900~1100 chars (with spaces). - Then, a blank line, then the same content, no subtitles/headings. - No subtitles/headings like "Korean Review", "StarCraft Broodwar Map Review", etc.

- Output starts immediately with the review content body. - Total output: 1800~3000 chars (so Korean + English + blank line). - Content should naturally cover: map structure, genre, gameplay style, play experience, recommended players, pre-play check items. - Do not invent maker, triggers, win conditions, story, balance numbers.

- Do not list map size, resolution, tile units, encoding, or repeat the same explanation. - Output must start immediately with the first sentence of the review body." - This is contradictory/confusing. The user prompt starts with "다음 를 자연스러운

로 번역해라..." and then has "Here's a thinking process:" which seems like standard AI prompt injection or a test. The actual map data is given after that? Let me check the structure. - The prompt seems to be: User message starts with "다음 를 자연스러운

로 번역해라..." then there's "Here's a thinking process:" and then a long block of instructions. Then map data is provided? Actually, looking at the end: there's map data: "Title: Untitled Scenario. Filename: 4P-Incomes-A.scx Internal title: Untitled Scenario. Internal description: 9527-7104=2423 (this looks like a code/math thing... Genre: MELEE (밀리) -> Melee.

Players: 0 Teams: - Playtime: - Difficulty: - Version: - Tags: - Tileset: Twilight (황혼)" - I need to figure out what exactly is expected. The prompt says: "다음 를 자연스러운

로 번역해라. 제목, 머리말, 목록, Markdown, 부연 설명 없이 한국어 번역 본문만 출력해라. 원문에 없는 맵 정보나 평가를 추가하지 마라." - Then there's the AI thinking block. Usually, in these tasks, the "thinking process" is for me to reason, and then I output the final answer. The final answer should be the Korean review translation adhering to the constraints.

- The constraints from the "thinking process" block are very specific: 900-110