生育救世主
8.4K posts

生育救世主
@Rollkey4
真正的赛博朋克。 命理师、独立开发者。 ACGN人格数据库:https://t.co/W2SMZvujqm 床搭子互助平台:https://t.co/6214wunOJX 工具站:https://t.co/ep28vQJevs 玄学命理博客站:https://t.co/KCrplLffXQ







使用agent-sprite-forge這個skill就可以在Codex上面直接做到one shot建立完整的遊戲了。 下面這2個全部都是在 Codex 使用GPT-5.5跟agent-sprite-forge "一次"做出來的,所有遊戲元素一次生成並且把遊戲邏輯都做出來給你了,後續哪裡有問題也是直接跟codex對話就行,不論是要加新的怪物還是新的任何元素都可以。 我還記得2年前要一張一張用Midjourney生成,然後慢慢丟到canva去背再慢慢寫code... 獨立遊戲開發生態,是真的要被 Agent 改寫了。



ChatGPT-Image-2.0 × Seedance 2.0 RPGゲームデモ 作り方: Step 1:ChatGPT-Image-2.0でゲーム風の画面を生成 例: Prompt ➕ image 「一人称」視点、「ARPGタイプ」、「西部劇系」ゲームに着想を得た実機風スクリーンショットを生成。NPCは参考画像1を使用。 セリフ: “It’s time to set out, adventurer.” 選択肢: Alright, let’s go. Not right now. Why not? 💡お好みの画風やゲームジャンルに置き換え Step 2:Seedance 2.0で静止画を動画化 テキストでUI操作、セリフ、キャラクターの動きなどを補足。 Prompt Share 120 First-person perspective, immersive ARPG dialogue scene, single continuous shot. A small white rabbit adventurer stands directly in front of the camera. Its fur is soft and finely detailed, and its right eye is a glowing red mechanical eye. It reaches out toward the camera in an inviting gesture, with a calm and confident demeanor. The environment is a vast fantasy world: rolling green grasslands, ruined ancient stone structures, distant mountain ranges, and a lightly smoking volcano. The morning light is soft, the air is clear, and a gentle breeze moves the grass and leaves, creating a strong cinematic atmosphere with depth. The camera is in first-person view, with slight handheld motion and natural breathing sway. Shallow depth of field, cinematic realism. The rabbit looks directly into the camera and says: “Good morning, adventurer. It’s time to set out.” A semi-transparent game UI (modern ARPG style) appears on the right side: Alright, let’s go. Not right now. Why not? The cursor moves smoothly twice, pauses for 0.3 seconds, then selects “Why not?” The rabbit smiles slightly, tilts its head, and responds in a relaxed tone: “Sure.” In the next moment, all UI elements (dialogue box, options, health bar, minimap) fade out smoothly, returning the scene to a clean, UI-free world. The rabbit turns around, its cloak and backpack swaying naturally with physics, and begins walking forward along a stone path. The camera maintains a first-person follow, with subtle handheld motion, synchronized with the character’s walking rhythm. While walking, the rabbit raises its hand and points toward the distant horizon (mountains and a faint city silhouette), saying: “Look, that’s our destination.” A brief pause of about 0.5 seconds. The wind intensifies slightly, causing more noticeable movement in the grass and clothing, enhancing the epic atmosphere. The rabbit continues walking and says: “We have to get there before 8:08 PM.” The camera slowly pushes forward, focusing toward the distance. The far view transitions from slightly blurred to clear, with a subtle lens flare appearing, reinforcing the sense of a journey beginning and a clear objective. No subtitles, no background music. Only natural ambient sounds: wind, footsteps, and distant birds. Tips: ChatGPT-Image-2.0のストーリーボードも便利なんですが、今回みたいな用途だとテキストで流れを指定した方が、全体のつながりは安定しやすい印象です。 逆に、カット単位でしっかり作り込みたい場合はストーリーボードの方が向いています。 用途次第で使い分けるのが良さそうです🐰 どんなジャンルのゲームが好きですか? 一緒に遊ぼう❤️🔥















