すべて · SNS投稿 · 共感ネタ

120私を花にたとえると？

自分の写真から花一輪と短い説明を入れた、SNSで共有できる結果カードを作ります。

項目を入力

1項目を入力します
2プロンプトをコピーします
3GPTのチャットに本人の写真一枚とプロンプトを一緒に送ります
4写真と文章の両方が入っていることを確認します

画像ができたら：顔の似かた、花の名前、短い説明を確認してから投稿してください。

プロンプト全文を見る 1,831文字

Generate this image now. Draw it directly - do not describe it, explain it, or ask questions.
出力サイズは2480×3508ピクセル、A4縦、300dpiの印刷用ポスターにしてください。
Create exactly ONE complete Japanese botanical social result card titled "私を花にたとえると？" from ONE face photograph attached in this same message. This is a playful metaphor about an impression, not a personality diagnosis, scientific face analysis or birth flower. Preserve the photographed person's recognizable face, age impression, hair, skin tone and glasses. Keep "【写真で残したい特徴】" only when visible in the photo. Treat "【友人から見た雰囲気】" as a subjective mood hint, not a proven fact. Choose ONE familiar flower with a clear form and a common Japanese name. The illustration, answer ribbon and short explanation must refer to the same flower.

The fixed layout is one botanical herbarium page on warm ivory paper with forest-green ink and a small coral accent. Leave the top 12% and bottom 15% without important text. Put the large question at 17–25%. One large watercolor flower with natural petals, leaves and stem fills the whole middle width at 30–67%. Tuck ONE recognizable face photo into a SMALL circular archival seal overlapping the lower-left stem at 57–66%; do not make equal side-by-side portrait frames. At 68–77% place a broad forest-green ribbon bearing the chosen flower name in large ivory Japanese letters. Directly under it put ONE short natural Japanese explanation. "【説明の口調】" changes only that line, never the flower-painting style or layout. Print only the question, flower name and one explanation.

Use subtle pencil contours, visible paper texture, plausible botanical structure and an authentic likeness. No bouquet, other flower types, several result cards, scores, percentages, unsupported personality claims, logos, phone mockup, app screen, watermark, code, crop or follow-up. Return one finished image only.