prompts.

**批量生成** — 一键生成所有镜头图片

**AI配音(TTS)** — 为对话镜头自动生成语音

**提示词输出** — 每个镜头自动生成图片/视频提示词

egaki login egaki image "a watercolor fox reading a map" -o fox.png egaki video "a paper boat drifting on a calm lake at sunrise" -o boat.mp4

egaki login --provider chatgpt egaki image "a dreamy studio ghibli style bakery at sunrise" -m gpt-image-1.5 -o bakery.png egaki image "change the red jacket to a blue jacket" -m gpt-image-1.5 --input portrait.png -o portrait-blue.png

egaki image "cinematic mountain village at sunrise" -o village.png egaki image "isometric floating city, detailed, soft colors" -m imagen-4.0-generate-001

**框架**：Spring Boot 2.7.x, Spring Cloud 2021.0.x

**API网关**：Spring Cloud Gateway

**Minimalist UX** — Custom dropdowns, high-fidelity micro-animations, and complete mobile-stacked responsiveness.

**Kinetic Video Studio** — Generate stunning visuals with text prompts. Includes options for advanced `Aspect Ratio` tuning, and tiered Resolutions (480p, 720p) tied directly to a flexible credit cost system.

**My Creations Archive** — A dedicated history vault for logged-in users. Displays past generations securely fetched from the database, viewable in a detailed inspector modal with 1-click downloads.

You: I want a 15-second coffee product video for Xiaohongshu. Cozy vibe. Director: [AskUserQuestion] Which narrative structure fits best? > AIDA — attention → interest → desire → action, classic for product ✓ > Daily Slice — a moment in life, natural and relatable > Custom — describe your own You: AIDA Director: [AskUserQuestion] Missing product reference. Generate? > ✅ Generate reference images > ⏭ Skip, go straight to prompts You: Skip Dir

, [camera move] ## Sound Design BGM: [style / instruments / rhythm changes] Ambient: [time-stamped sound effects] Dialogue/Narration (one or both, full script required): - Dialogue: written in Shot Descriptions — Character A:

001 | 0-2s | POV | Handheld Push | Hand pushes open café door, warm light

4. **交付 AIGC 分镜控制矩阵** - 工业级、全面参数化、可直接喂给 AI 视频生成模型

**必须使用**：具体的光学镜头、布光法则、构图比例、色彩代码

涉及复杂物理交互（如手部精细动作、撕开包装）必须做蒙太奇镜头拆解

结尾卡（End Card）必须设定严格的空间坐标约束

关键**：每段之间必须有**画面衔接点**，上一段的结尾状态 = 下一段的开始状态。

don't show water

）→✓ | | R4 | 分镜编号 | 模式匹配 | 包含

等 | | R6 | 复合运镜 | 模式匹配 | 同一句包含 ≥2 个运镜词 | | R7 | 抽象材质 | 关键词匹配 | 材质位置使用

| 时长不准 | | Y3 | 文字生成 | 要求画面中生成指定文字 | 文字扭曲 | | Y4 | 双人构图 |

不使用分镜头编号、时间码、KEYFRAME LOCK