Back to Details

movie-maker

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Movie Maker Skill

影片制作Skill

This repo is docs-first: it ships guidance (rules + schemas), not an executable pipeline. This is a skill created with LongStories.ai Skill SEO Audit.
本仓库以文档为核心：它提供指导内容（规则+Schema），而非可执行的流水线。 这是一款通过LongStories.ai Skill SEO Audit创建的Skill。

When to use this skill

何时使用本Skill

Use this skill when a user wants a 10-60s mini-movie with a script, shotlist, and render-ready manifest.
当用户需要一个包含剧本、镜头清单且可直接用于渲染的清单文件的10-60秒迷你影片时，可使用本Skill。

What this skill does NOT do

本Skill不具备的功能

  • Does not include a runnable pipeline
  • Does not store or upload media
  • Does not include provider keys or secrets
  • 不包含可运行的流水线
  • 不存储或上传媒体文件
  • 不包含服务提供商的密钥或机密信息

What the agent should produce

Agent应生成的内容

  1. A numbered script (lines 
    L001…
    )
  2. A shotlist divided into scenes (
    S01…
    ), referencing script lines
  3. A manifest JSON that combines both and is ready to render
  1. 编号剧本（格式为
    L001…
    ）
  2. 镜头清单：按场景划分（格式为
    S01…
    ），并关联剧本对应台词
  3. 清单JSON文件：整合剧本与镜头清单，可直接用于渲染

Start here (MUST do intake)

入门必读（必须完成信息收集）

Read and follow: 
rules/start/start-here.md
Full runbook: 
rules/start/end-to-end.md
阅读并遵循：
rules/start/start-here.md
完整操作手册：
rules/start/end-to-end.md

Intake (copy/paste)

信息收集（可直接复制使用）

Ask these in one message (use an "ask the user questions" tool if available):
  1. Length: 30s / 45s / 60s?
  2. Aspect ratio: 16:9 or 9:16?
  3. Voice: narrator-only (default) or dialogue?
  4. Characters: names + 1-2 sentence vibe each (we will lock a canonical descriptor)
  5. Music: none (default) or background music?
  6. Style: pick a preset (cinematic / 3D animated / anime / 2D / photoreal)
  7. Quality mode: Draft (480p), Standard (720p), HQ (1080p)
  8. Model preset: Cheap and Fast (Z-Image Turbo + Seedance v1 fast) or High Quality, Expensive (Seedream 4.5 + Seedance 1.5 pro)
  9. Reference images: any character/style refs? (if yes, ask user to paste/attach and say you will save to 
    assets/refs/...
    )
If the user does not answer, propose defaults and ask:
"If I don't hear back, I'll do 30s, 16:9, narrator-only, 3D animated style, Draft (480p), High Quality, Expensive preset. Confirm?"
在一条消息中询问以下问题（如果有“向用户提问”工具可直接调用）：
  1. 时长：30秒 / 45秒 / 60秒？
  2. 宽高比：16:9 或 9:16？
  3. 语音形式：仅旁白（默认）或对话形式？
  4. 角色：角色名称 + 1-2句话的风格描述（我们将确定标准化的描述内容）
  5. 音乐：无（默认）或添加背景音乐？
  6. 风格：选择预设风格（电影质感 / 3D动画 / 动漫 / 2D平面 / 照片级写实）
  7. 质量模式：草稿级（480p）、标准级（720p）、高清级（1080p）
  8. 模型预设：低成本快速型（Z-Image Turbo + Seedance v1 fast）或高质量高成本型（Seedream 4.5 + Seedance 1.5 pro）
  9. 参考图片：是否有角色/风格参考图？（如果有，请让用户粘贴/上传，并告知将保存至
    assets/refs/...
    路径）
如果用户未回复，可提出默认选项并确认：
“若未收到回复，我将按照30秒、16:9比例、仅旁白、3D动画风格、草稿级（480p）、高质量高成本预设来生成。是否确认？”

Core pipeline (voiceover-first)

核心流水线（以旁白为核心）

text
Script -> Voiceover timing -> Finalize shot durations -> Generate images -> Generate videos -> Render locally -> Share
text
剧本 -> 旁白时长规划 -> 最终确定镜头时长 -> 生成图片 -> 生成视频 -> 本地渲染 -> 分享

Model presets (image + video combo)

模型预设（图像+视频组合）

See: 
rules/providers/model-presets.md
Quick summary:
  • Cheap and Fast: Z-Image Turbo + Seedance v1 fast
  • High Quality, Expensive: Seedream 4.5 + Seedance 1.5 pro
详情见：
rules/providers/model-presets.md
快速摘要：
  • 低成本快速型：Z-Image Turbo + Seedance v1 fast
  • 高质量高成本型：Seedream 4.5 + Seedance 1.5 pro

Pricing verification

价格验证

Prices change. Use the fal pricing endpoint in:
  • rules/providers/pricing.md
价格可能变动，请使用fal定价接口，详情见：
  • rules/providers/pricing.md

Rules index

规则索引

Ordered discovery path: 
rules/index.md
Provider-specific guidance lives under 
rules/providers/
. All other rules should remain provider-agnostic, with defaults listed in 
rules/providers/defaults.md
.
推荐浏览顺序：
rules/index.md
服务提供商专属指导内容位于
rules/providers/
路径下。所有其他规则应保持与提供商无关，默认配置见
rules/providers/defaults.md
。

Start

入门

  • rules/start/start-here.md
  • rules/start/requirements.md
  • rules/start/pipeline-voiceover-first.md
  • rules/start/end-to-end.md
  • rules/start/start-here.md
  • rules/start/requirements.md
  • rules/start/pipeline-voiceover-first.md
  • rules/start/end-to-end.md

Planning (script to shotlist)

规划（从剧本到镜头清单）

  • rules/planning/script-writing.md
  • rules/planning/word-count.md
  • rules/planning/scenes.md
  • rules/planning/shots.md
  • rules/planning/quality-and-cost.md
  • rules/planning/script-writing.md
  • rules/planning/word-count.md
  • rules/planning/scenes.md
  • rules/planning/shots.md
  • rules/planning/quality-and-cost.md

Manifest

清单文件

  • rules/manifest/schema.md
  • rules/manifest/script-and-shotlist.md
  • rules/manifest/schema.md
  • rules/manifest/script-and-shotlist.md

Audio (provider-agnostic)

音频（与提供商无关）

  • rules/audio/pauses-and-roomtone.md
  • rules/audio/captions-srt.md
  • rules/audio/pauses-and-roomtone.md
  • rules/audio/captions-srt.md

Visuals (provider-agnostic)

视觉效果（与提供商无关）

  • rules/visuals/image-prompting.md
  • rules/visuals/video-prompting.md
  • rules/visuals/style-presets.md
  • rules/visuals/characters/character-descriptors.md
  • rules/visuals/characters/consistency.md
  • rules/visuals/image-prompting.md
  • rules/visuals/video-prompting.md
  • rules/visuals/style-presets.md
  • rules/visuals/characters/character-descriptors.md
  • rules/visuals/characters/consistency.md

Timing

时长规划

  • rules/timing/from-timestamps.md
  • rules/timing/map-lines-to-timestamps.md
  • rules/timing/clamp-and-trim.md
  • rules/timing/from-timestamps.md
  • rules/timing/map-lines-to-timestamps.md
  • rules/timing/clamp-and-trim.md

Providers

服务提供商

  • rules/providers/defaults.md
  • rules/providers/model-presets.md
  • rules/providers/fal/overview.md
  • rules/providers/fal/api.md
  • rules/providers/elevenlabs/voiceover.md
  • rules/providers/fal/voiceover-elevenlabs.md (makes ElevenLabs optional)
  • rules/providers/elevenlabs/transcription.md
  • rules/providers/elevenlabs/music.md
  • rules/providers/defaults.md
  • rules/providers/model-presets.md
  • rules/providers/fal/overview.md
  • rules/providers/fal/api.md
  • rules/providers/elevenlabs/voiceover.md
  • rules/providers/fal/voiceover-elevenlabs.md（使ElevenLabs成为可选服务）
  • rules/providers/elevenlabs/transcription.md
  • rules/providers/elevenlabs/music.md

Rendering

渲染

  • rules/render/rendering.md
  • rules/render/assets.md
  • rules/render/rendering.md
  • rules/render/assets.md

QA + Troubleshooting

质量检查与故障排除

  • rules/qa/image-review.md
  • rules/qa/video-spotcheck.md
  • rules/troubleshooting/common-errors.md
  • rules/qa/image-review.md
  • rules/qa/video-spotcheck.md
  • rules/troubleshooting/common-errors.md

Sharing

分享

  • rules/sharing/clawtube.md
  • rules/sharing/clawtube.md

Examples

示例

  • examples/shipped/
    (script + shotlist + manifest)
  • examples/shipped/
    （包含剧本+镜头清单+清单文件）

Advanced

进阶内容

  • rules/advanced/overrides.md (per-scene/per-shot model overrides)
  • rules/advanced/overrides.md（按场景/镜头覆盖模型配置）

Next rules to add

待添加规则

  • Provider/model matrix docs
  • Safety: secrets, prompt injection, and content policies
  • 服务提供商/模型矩阵文档
  • 安全相关：机密信息、提示词注入、内容政策