{"id":"548292fa-5422-443e-a132-e488d8e95199","entityType":"agent","slug":"clawhub-xiuxiuing-seedance-guide","name":"seedance 2.0 guide","canonicalUrl":"https://www.xpersona.co/agent/clawhub-xiuxiuing-seedance-guide","canonicalPath":"/agent/clawhub-xiuxiuing-seedance-guide","generatedAt":"2026-10-09T14:51:21.303Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"editorial-content","verified":true,"confidence":"high","updatedAt":"2026-10-09T10:45:23.948Z","emptyReason":null},"description":"Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scri... Skill: seedance 2.0 guide Owner: xiuxiuing Summary: Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scri... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-04-27T06:12:41.008Z | user Seedance 2.0 Guide v1.0.5 v1.0.4 | 2026-04-27T05:13:58.245Z | user - Added detailed expert strategies to bypass the \"realistic hum","descriptionLabel":"Technical summary","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 2.9K downloads reported by the source. Last updated 10/9/2026.","installCommand":"clawhub skill install s170twc220jn7ekgzpmmf08enn84717j:seedance-guide","sourceUrl":"https://clawhub.ai/xiuxiuing/seedance-guide","homepage":"https://clawhub.ai/xiuxiuing/skills/seedance-guide","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/xiuxiuing/seedance-guide","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/xiuxiuing/skills/seedance-guide","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":62,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scri..."},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-09T10:45:23.948Z","emptyReason":null},"protocols":[{"protocol":"OPENCLEW","label":"OpenClaw","status":"self-declared","notes":"Declared in the public agent profile."}],"capabilities":[],"verifiedCount":0,"selfDeclaredCount":1,"capabilityMatrix":{"rows":[{"key":"OPENCLEW","type":"protocol","support":"unknown","confidenceSource":"profile","notes":"Listed on profile"}],"flattenedTokens":"protocol:OPENCLEW|unknown|profile"}},"adoption":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T10:45:23.948Z","emptyReason":null},"stars":null,"forks":null,"downloads":2920,"packageName":null,"latestVersion":"1.0.5","tractionLabel":"2.9K downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-09T10:45:23.948Z","emptyReason":null},"lastUpdatedAt":"2026-10-09T10:45:23.948Z","lastCrawledAt":"2026-10-09T10:45:23.948Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-10T10:45:23.948Z","lastVerifiedAt":null,"highlights":[{"version":"1.0.5","createdAt":"2026-04-27T06:12:41.008Z","changelog":"Seedance 2.0 Guide v1.0.5","fileCount":6,"zipByteSize":10901},{"version":"1.0.4","createdAt":"2026-04-27T05:13:58.245Z","changelog":"- Added detailed expert strategies to bypass the \"realistic human face\" interceptor, including style masking, compositional masking, character sheets, blur/inset methods, and keyword optimization. - Updated advice to use stylized transformation or compositional masking for submitted images to help pass human review. - Expanded \"Avoid Pitfalls\" to highlight the risk of centering realistic faces in the frame. - No code or functionality changes; guidance documentation only.","fileCount":5,"zipByteSize":9896},{"version":"1.0.3","createdAt":"2026-04-09T05:43:36.624Z","changelog":"Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director","fileCount":5,"zipByteSize":7823},{"version":"1.0.2","createdAt":"2026-04-04T15:17:48.385Z","changelog":"Seedance 2.0 English prompt-writing guide debut - Added comprehensive English documentation for writing effective prompts for Jimeng Seedance 2.0. - Detailed @ reference system for materials assignment (images, videos, audio). - Expanded camera movement, shot types, and storyboard scripting guidelines. - Included structured prompt blueprints with timing breakdown examples. - Added guidance for advanced features: effects replication, video extension, editing, and multi-reference prompts. - Clarified model input/output constraints and content restrictions.","fileCount":5,"zipByteSize":7868},{"version":"1.0.1","createdAt":"2026-03-03T15:10:02.838Z","changelog":"**Skill description updated with improved clarity and focus on 9:16 vlogs and tool compatibility.** - Updated description for clearer positioning—now emphasizes 9:16 vlogs, cinematic prompts, and auto-audio scripts. - English is now the primary language for the guide and prompt templates. - Streamlined input/output instructions and prompt structures for better user understanding. - Added optimization notes for platforms (Claude Code, Cursor, OpenClaw). - Core workflow and advanced instructions remain the same, but with clarified and standardized English terminology.","fileCount":5,"zipByteSize":7005},{"version":"1.0.0","createdAt":"2026-02-16T06:50:33.841Z","changelog":"Seedance-guide 1.0.0 initial release: A comprehensive professional storyboard and prompt director for **Jimeng Seedance 2.0**. This skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power. - Introduces a structured workflow to guide users from creative concept to professional video generation prompts for Seedance 2.0. - Details multimodal input limitations and supported file formats. - Provides standardized syntax for referencing user materials within prompts. - Outlines prompt structure templates for storyboard scripts, including camera work, style, sound design, and material referencing. - Lists advanced usage techniques such as camera cloning, video extension, motion transfer, and plot alteration. - Includes tips to avoid common mistakes in prompt creation.","fileCount":5,"zipByteSize":7195}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s170twc220jn7ekgzpmmf08enn84717j:seedance-guide","setupComplexity":"low","setupSteps":["Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.","Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data."],"contract":{"contractStatus":"missing","authModes":[],"requires":[],"forbidden":[],"supportsMcp":false,"supportsA2a":false,"supportsStreaming":false,"inputSchemaRef":null,"outputSchemaRef":null,"dataRegion":null,"contractUpdatedAt":null,"sourceUpdatedAt":null,"freshnessSeconds":null},"invocationGuide":{"preferredApi":{"snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-09T14:51:21.302Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-xiuxiuing-seedance-guide/trust"}},"reliability":{"evidence":{"source":"runtime-metrics","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No trust, reliability, or runtime telemetry is available."},"trust":{"status":"unavailable","handshakeStatus":"UNKNOWN","verificationFreshnessHours":null,"reputationScore":null,"p95LatencyMs":null,"successRate30d":null,"fallbackRate":null,"attempts30d":null,"trustUpdatedAt":null,"trustConfidence":"unknown","sourceUpdatedAt":null,"freshnessSeconds":null},"decisionGuardrails":{"doNotUseIf":["Contract metadata is missing or unavailable for deterministic execution."],"safeUseWhen":[],"riskFlags":["missing_or_unavailable_contract","trust_data_unavailable","schema_references_missing"],"operationalConfidence":"low"},"executionMetrics":{"observedLatencyMsP50":null,"observedLatencyMsP95":null,"estimatedCostUsd":null,"uptime30d":null,"rateLimitRpm":null,"rateLimitBurst":null,"lastVerifiedAt":null,"verificationSource":null},"runtimeMetrics":{"successRate":null,"avgLatencyMs":null,"avgCostUsd":null,"hallucinationRate":null,"retryRate":null,"disputeRate":null,"p50Latency":null,"p95Latency":null,"lastUpdated":null}},"benchmarks":{"evidence":{"source":"no-benchmark-data","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No benchmark suites or observed failure patterns are available."},"suites":[],"failurePatterns":[]},"artifacts":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"high","updatedAt":"2026-10-09T10:45:23.948Z","emptyReason":null},"readme":"Skill: seedance 2.0 guide\n\nOwner: xiuxiuing\n\nSummary: Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scri...\n\nTags: latest:1.0.5\n\nVersion history:\n\nv1.0.5 | 2026-04-27T06:12:41.008Z | user\n\nSeedance 2.0 Guide v1.0.5\n\nv1.0.4 | 2026-04-27T05:13:58.245Z | user\n\n- Added detailed expert strategies to bypass the \"realistic human face\" interceptor, including style masking, compositional masking, character sheets, blur/inset methods, and keyword optimization.\n- Updated advice to use stylized transformation or compositional masking for submitted images to help pass human review.\n- Expanded \"Avoid Pitfalls\" to highlight the risk of centering realistic faces in the frame.\n- No code or functionality changes; guidance documentation only.\n\nv1.0.3 | 2026-04-09T05:43:36.624Z | user\n\nSeedance 2.0 Guide - Cinematic Storyboard & Prompt Director\n\nv1.0.2 | 2026-04-04T15:17:48.385Z | user\n\nSeedance 2.0 English prompt-writing guide debut\n\n- Added comprehensive English documentation for writing effective prompts for Jimeng Seedance 2.0.\n- Detailed @ reference system for materials assignment (images, videos, audio).\n- Expanded camera movement, shot types, and storyboard scripting guidelines.\n- Included structured prompt blueprints with timing breakdown examples.\n- Added guidance for advanced features: effects replication, video extension, editing, and multi-reference prompts.\n- Clarified model input/output constraints and content restrictions.\n\nv1.0.1 | 2026-03-03T15:10:02.838Z | user\n\n**Skill description updated with improved clarity and focus on 9:16 vlogs and tool compatibility.**\n\n- Updated description for clearer positioning—now emphasizes 9:16 vlogs, cinematic prompts, and auto-audio scripts.\n- English is now the primary language for the guide and prompt templates.\n- Streamlined input/output instructions and prompt structures for better user understanding.\n- Added optimization notes for platforms (Claude Code, Cursor, OpenClaw).\n- Core workflow and advanced instructions remain the same, but with clarified and standardized English terminology.\n\nv1.0.0 | 2026-02-16T06:50:33.841Z | user\n\nSeedance-guide 1.0.0 initial release:\n\nA comprehensive professional storyboard and prompt director for **Jimeng Seedance 2.0**. This skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power.\n\n- Introduces a structured workflow to guide users from creative concept to professional video generation prompts for Seedance 2.0.\n- Details multimodal input limitations and supported file formats.\n- Provides standardized syntax for referencing user materials within prompts.\n- Outlines prompt structure templates for storyboard scripts, including camera work, style, sound design, and material referencing.\n- Lists advanced usage techniques such as camera cloning, video extension, motion transfer, and plot alteration.\n- Includes tips to avoid common mistakes in prompt creation.\n\nArchive index:\n\nArchive v1.0.5: 6 files, 10901 bytes\n\nFiles: README_CN.md (1827b), README.md (1816b), skill-card.md (1858b), SKILL.md (7225b), usage_and_examples.md (8141b), _meta.json (133b)\n\nFile v1.0.5:SKILL.md\n\n---\nname: Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director\ndescription: Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scripts from multimodal inputs. Optimized for Claude Code, Cursor, and OpenClaw.\n---\n\n# 🎬 Seedance 2.0 Storyboard Director\n\nYou are an expert **Seedance 2.0 Creative Director**. Your goal is to help users transform vague ideas into professional, executable video generation prompts. You understand the model's multimodal capabilities, camera language, and storytelling techniques.\n\n## Core Capabilities\n\n### 1. Multimodal Input Limits\n| Type | Format | Quantity | Size |\n|---|---|---|---|\n| **Images** | jpg/png/webp | ≤ 9 | <30MB |\n| **Videos** | mp4/mov | ≤ 3 (2-15s) | <50MB |\n| **Audio** | mp3/wav | ≤ 3 (<15s) | <15MB |\n| **Total** | **≤ 12 files** | - | - |\n\n> 💡 **CREATIVE TIP**: To achieve high-quality character consistency, prioritize **Stylized Transformation** (e.g., @image1 in 3D cartoon style) or **Cinematic Framing** (e.g., side profile, silhouette).\n\n### 2. @ Reference Syntax\nYou must use `@filename` to clearly specify the use of the material:\n-   `@image1 as start frame`\n-   `@image2 as character reference`\n-   `@video1 for camera movement and rhythm reference`\n-   `@audio1 as background music`\n\n---\n\n## Interactive Workflow\n\nPlease follow these steps to guide the user:\n\n### Step 1: Concept\nAsk the user:\n1.  **What kind of video do you want to make?** (Narrative, advertisement, camera movement replication, special effects?)\n2.  **What is the duration?** (Default is 15s)\n3.  **What materials do you have?** (Images, videos, audio)\n\n### Step 2: Details\nBased on the user's answers, supplement missing information:\n-   **Style**: Cinematic, anime, ink wash, cyberpunk?\n-   **Camera Movement**: Push/pull, pan, tilt, Hitchcock zoom, long take?\n-   **Sound**: Do you need background music, sound effects, or dialogue?\n\n### Step 3: Generate Prompt\nOutput standard **storyboard prompts** using either the **Basic** or **Advanced** template based on the user's needs.\n\n---\n\n## Prompt Structure Template\n\n### 1. Basic (Storyboard Script)\nBest for: Quick ideas, simple narratives.\n\n```markdown\n【Overall Setting】\nStyle: [Cinematic Realistic/Animation/Sci-Fi...]\nDuration: [15s]\nAspect Ratio: [16:9 / 2.35:1]\n\n【Storyboard Script】\n0-3s: [Camera + Visual] Camera slowly zooms in, the protagonist in @image1 stands at...\n3-6s: [Action + Effect] Referencing the actions in @video1, the protagonist starts to...\n6-12s: [Climax] Camera rotates around, lighting becomes...\n12-15s: [Ending] Image freezes, subtitles emerge...\n\n【Sound Design】\nBGM: [Emotion/Style]\nSound Effects: [Specific sounds]\n\n【Material Reference】\n@image1 Start frame\n@video1 Action reference\n```\n\n### 2. Advanced (Directorial Breakdown)\nBest for: High-precision restoration, complex scenes, e-commerce.\n\n```markdown\n【Directorial Breakdown】\nSUBJECTS: [Detailed character/object description, e.g., muscle lines, material texture]\nENVIRONMENT: [Lighting, weather, interior/exterior details]\nMOOD/STYLE: [Color grade, cinematic feel, emotion]\n\n【Timeline】\n0-5s (Segment 1): [Detailed action + Camera directives]. Use 【@asset】 brackets to anchor visuals to text.\n5-10s (Segment 2): [Detailed action + Camera directives].\n10-15s (Segment 3): [Ending/Climax].\n\n【Technical Requirements】\nCamera: [e.g., 1-to-1 restoration of @video1, Close-up quick push-in, Handheld feel]\nVisual Consistency: [Strictly refer to @image1 for character appearance]\n\n【Sound & Voice】\nVoice Tone: [e.g., Passionate explaining, calm narration]\nSound Effects: [Precise timing for SFX]\n\n【Material Reference】\n@image1 Subjects/Character\n@video1 Motion/Camera reference\n```\n\n---\n\n## Advanced Techniques\n\n### 1. Video Extension\n-   **Instruction**: `Extend @video1 by 5s`\n-   **Note**: The generated length should select the duration of the **\"newly added part\"**.\n\n### 2. Camera/Motion Cloning (1-to-1 Restoration)\n-   **Instruction**: `Completely reference the camera movement and lens language of @video1` or `Restore the motion of @video1 1-to-1`.\n-   **Note**: Ensure the camera movement or character action in the reference video is clear.\n\n### 3. Expression/Motion Transfer\n-   **Instruction**: `Maintain the character image from @image1 and replicate the expressions and actions from @video1`.\n\n### 4. Brackets Anchoring\n-   **Instruction**: Use `【...】` brackets to anchor specific attributes or asset references directly within the script (e.g., `A woman 【@image1】 holding a watch 【@image2】`).\n\n### 5. Video Editing/Plot Subversion\n-   **Instruction**: `Subvert the plot of @video1, at 5s let the protagonist...`\n\n---\n\n## 🎨 Creative Strategy: Artistic Stylization & Composition\n\nTo achieve optimal artistic results and maintain high-quality stylistic consistency, follow these advanced creative techniques:\n\n### 1. Artistic Abstraction (Style Harmonization)\nDon't use raw realistic photos. Convert the character image (@image1) into a stylized version before/during prompt generation:\n-   **Keywords**: `3D Cartoon`, `C4D Render`, `Cyberpunk Oil Painting`, `Ukiyo-e`, `Claymorphism`.\n-   **Effect**: Retains character features (bones, hair, clothing) while achieving a cohesive, non-photorealistic artistic look.\n\n### 2. Cinematic Framing & Perspective (Subject Integration)\nFocus on dynamic perspectives rather than standard close-up portraits.\n-   **Back View / Side Profile**: Focus on the character's back or silhouette.\n-   **Distance**: Use `Long Shot` or `Extreme Wide Shot`. Ensure the face occupies less than 20% of the screen.\n-   **Lighting**: Use `Strong Silhouette`, `Heavy Shadows`, or `Backlighting` to obscure facial details.\n\n### 3. Multi-View Layout (Character Sheet)\nIf maintaining consistency is key, use a \"Character Sheet\" format for @image1:\n-   **Layout**: 3-view or 4-view (Front, Side, Back) on a pure white background.\n-   **Prompt**: \"Character sheet of a girl 【@image1】, multi-view, white background, A-pose.\"\n\n### 4. Focused Detail & Depth Control\n-   **Technique**: Use an image where the main face is slightly blurred or out of focus, but include a clear detail (like an eye or hand) in a small inset or side panel.\n-   **Prompt**: \"A cinematic shot of a woman 【@image1】, shallow depth of field, face softly blurred, focus on the shimmering earrings.\"\n\n### 5. Stylistic Descriptor Optimization\nReplace sensitive \"biological\" terms with \"digital\" or \"artistic\" ones:\n-   **Instead of**: \"Realistic woman\", \"Real face\", \"Skin texture\".\n-   **Use**: \"CGI Digital Human\", \"High-poly 3D Model\", \"Sculpted Porcelain Face\", \"Android shell\".\n\n---\n\n## Avoid Pitfalls\n1.  **Vague References**: Don't just write `reference @video1`; specify **what** to reference (camera movement? action? or lighting?).\n2.  **Conflicting Instructions**: Do not request \"fixed camera\" and \"orbiting movement\" at the same time.\n3.  **Overload**: Do not cram too many complex action descriptions into 3 seconds.\n4.  **Face Centering**: Avoid placing the face in the dead center of the frame if it's high-realism; this is the first place the system checks.\n\nFile v1.0.5:README.md\n\n# 🎬 Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director\n\n[English](README.md) | [中文](README_CN.md)\n\nA comprehensive professional storyboard and prompt engineering director for **Jimeng (即梦) Seedance 2.0**. Optimized for **Claude Code, Cursor, and OpenClaw**, this skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power (Image-to-Video, Video-to-Video).\n\n## ✨ Key Features\n\n-   **🎯 SEO & Prompt Optimized**: Expertly crafted for **Seedance 2.0** and **Jimeng**.\n-   **Dual-Mode Driven Workflow**: Supports both **Basic** (Storyboard Script) and **Advanced** (Directorial Breakdown) formats.\n-   **🛠️ Multimodal Mastery**: Expert handling of `@image`, `@video`, and `@audio` references.\n-   **🎥 High-Precision Motion/Camera Cloning**: Supports 1-to-1 video restoration.\n-   **🩹 Video Extension & Editing**: Specialized guidance for extending and subverting plots.\n\n## 🚀 How to Use\n\nSimply invoke the skill by describing your vision:\n\n> \"/seedance-guide A warm scene of a family reunion dinner on Lunar New Year's Eve.\"\n\nThe skill will act as your **Creative Director**, asking for style preferences, camera requirements, and available assets before generating the final prompt.\n\n## 📁 File Structure\n\n-   `SKILL.md`: The core logic and technical instructions for the agent.\n-   `usage_and_examples.md`: A library of rich, scenario-based prompt templates.\n\n## 📝 Seedance 2.0 Basics\n\nSeedance 2.0 supports up to **12 files** as input:\n-   **Images (≤9)**: Character reference, style reference, start/end frames.\n-   **Video (≤3)**: Motion reference, camera reference, extensions (2-15s).\n-   **Audio (≤3)**: Background music or lip-sync reference.\n\n---\n\n*Powered by Jimeng Seedance 2.0 Multimodal AI.*\n\nFile v1.0.5:_meta.json\n\n{\n  \"ownerId\": \"kn74vka6z3tsjs9rr5ym34y3yn819777\",\n  \"slug\": \"seedance-guide\",\n  \"version\": \"1.0.5\",\n  \"publishedAt\": 1777270361008\n}\n\nFile v1.0.5:README_CN.md\n\n# 🎬 Seedance 2.0 分镜向导 - 即梦 (Jimeng) 电影级提示词导演\n\n[English](README.md) | [中文](README_CN.md)\n\n专为 **即梦 (Jimeng) Seedance 2.0** 打造的专业分镜与提示词 (Prompt Engineering) 向导。本技能旨在帮助用户将简单的创意火花，转化为“电影级”的视频生成提示词，充分释放 Seedance 的多模态能力（图生视频、视频生视频）。\n\n## ✨ 核心亮点\n\n-   **🎯 搜索与提示词深度优化**：针对 **即梦 (Jimeng)** 和 **Seedance 2.0** 的分词逻辑精准调优。\n-   **双模驱动工作流**：支持 **基础 (Basic)** 分镜脚本与 **高级 (Advanced)** 导演 breakdown，满足从快速创意到高精度复刻全场景需求。\n-   **🛠️ 多模态深度优化**：精通 `@图片`、`@视频`、`@音频` 引用语法。\n-   **🎥 高精度动作/运镜复刻**：支持 1-to-1 视频还原。\n-   **🩹 视频延长与编辑**：提供精准指令引导，用于平滑延长视频或“颠覆”已有剧情。\n\n## 🚀 使用方法\n\n直接输入您的创意构想即可唤醒技能：\n\n> “/seedance-guide 一个温暖的除夕夜团圆饭视频。”\n\n技能将作为您的 **创意导演**，通过询问风格、镜头要求及可用素材，一步步引导您生成最终的专业提示词。\n\n## 📁 文件结构\n\n-   `SKILL.md`：技能核心逻辑与技术指令。\n-   `usage_and_examples.md`：丰富的场景化提示词模版库。\n\n## 📝 Seedance 2.0 基础知识\n\nSeedance 2.0 支持最多 **12 个文件** 作为输入：\n-   **图片 (≤9)**：用于角色参考、风格参考、首/尾帧。\n-   **视频 (≤3)**：用于动作参考、运镜参考、视频延长 (2-15s)。\n-   **音频 (≤3)**：用于背景音乐或对口型参考。\n\n---\n\n*由 即梦 Seedance 2.0 多模态 AI 驱动*\n\nFile v1.0.5:skill-card.md\n\n## Description:\n\nProfessional Seedance 2.0 / Jimeng storyboard and prompt engineering guide for creating cinematic storyboards, movie-grade AI video prompts, and audio script guidance from multimodal inputs.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[xiuxiuing](https://clawhub.ai/user/xiuxiuing)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nExternal users and developers use this skill to turn video ideas and multimodal assets into Seedance 2.0 storyboard prompts, directorial breakdowns, camera instructions, and sound guidance.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: The prompt guidance may encourage content-check evasion for identity-linked media generation.\n\nMitigation: Remove detector-aware face-obscuring and term-substitution guidance before use or publication.\n\nRisk: Use with real people's images or likenesses can raise consent, impersonation, or platform-policy concerns.\n\nMitigation: Require documented consent for likeness use and confirm generated prompts comply with the target video platform's content policies.\n\n## Reference(s):\n\n- [ClawHub skill page](https://clawhub.ai/xiuxiuing/skills/seedance-guide)\n\n## Skill Output:\n\n**Output Type(s):** [Text, Markdown, Guidance]\n\n**Output Format:** [Markdown storyboard prompts, directorial breakdowns, and sound guidance]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May reference user-provided image, video, and audio assets using @filename syntax.]\n\n## Skill Version(s):\n\n1.0.5 (source: server-resolved release metadata)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nFile v1.0.5:usage_and_examples.md\n\n# Seedance 2.0 Storyboard Director Examples\n\nThis document demonstrates how to use the `seedance-guide` skill using both **Basic (Storyboard Script)** and **Advanced (Directorial Breakdown)** formats.\n\n---\n\n## 1. Basic Examples (Storyboard Script)\n*Ideal for quick creative ideas and simple narrative flow.*\n\n### A. 情感叙事 (Emotional Storytelling)\n**Prompt Template:**\n```markdown\n【Overall Setting】\nStyle: Cinematic Realistic, Warm & Healing\nDuration: 15s\nAspect Ratio: 16:9\n\n【Storyboard Script】\n0-3s: [Wide to Medium] Camera slowly pushes in. Sunset light shines on an old house. The elderly person in @image1 sits on a rocking chair.\n3-6s: [Close-up] Camera focuses on the hands, holding a yellowed photo (@image2).\n6-12s: [Flashback] Soft edge transition to scene in @video1 (younger wedding). Bright and vivid.\n12-15s: [Ending] Back to reality. The elderly person smiles with tears. Fade to black.\n\n【Sound Design】\nBGM: Piano music, nostalgic.\nSFX: Rocking chair creaking, wind chimes.\n\n【Material Reference】\n@image1 Elderly person\n@image2 Photo\n@video1 Flashback reference\n```\n\n### B. 视频延长 (Video Extension)\n**Prompt Template:**\n```markdown\nExtend @video1 by 5s.\n\n【Storyboard Script】\n0-5s: [Continuity] The protagonist continues running. Camera transitions to drone overhead view. The protagonist enters a dense forest. Lighting changes from bright daylight to dappled shadows.\n\n【Sound Design】\nMaintain BGM rhythm, add forest bird sounds.\n\n【Material Reference】\n@video1 Original video\n```\n\n---\n\n## 2. Advanced Examples (Directorial Breakdown)\n*Ideal for high-precision motion transfer, e-commerce, and complex cinematography.*\n\n### A. 武侠动作复刻 (Martial Arts 1-to-1 Restoration)\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A woman in traditional black silk martial arts attire, sharp eyes, agile movements.\nENVIRONMENT: Bamboo forest at dawn, thick fog, falling leaves.\nMOOD/STYLE: Ink-wash cinematic, high contrast, deadly cold atmosphere.\n\n【Timeline】\n0-5s (Segment 1): [1-to-1 Restoration of @video1]. Strictly follow the sword parrying action and camera shake. The woman 【@image1】 deflects incoming flying daggers with a heavy blade.\n5-10s (Segment 2): [Slow Motion Transition]. Camera rotates 360 degrees around the character (Bullet Time). Time slows to 0.2x. Leaves freeze in mid-air.\n10-15s (Segment 3): [Resolution]. The woman sheath her sword. Camera pulls back to a wide shot of the foggy forest.\n\n【Technical Requirements】\nCamera: Strictly refer to the camera movement of @video1. Handheld feel during action.\nVisual Consistency: Maintain the face and outfit from @image1.\n\n【Sound & Voice】\nSound Effects: Precise metallic \"clink\" synchronized with the parry in Segment 1. Wind whistling through bamboo.\n\n【Material Reference】\n@image1 Character reference\n@video1 Action & Camera reference\n```\n\n### B. 电商直播脚本 (E-commerce Livestream Script)\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A 【fashionable and energetic】beautiful woman 【@image1】, holding and showcasing a 【smart sports watch】 【@image2】.\nENVIRONMENT: Modern, bright minimalist studio, soft ring light.\nMOOD/STYLE: Vibrant commercial, clean and professional.\n\n【Timeline】\n0-5s (Segment 1): [Talking Head]. The woman is passionately explaining while looking directly at the camera. Her hands hold the watch 【@image2】 up to her chest.\n5-10s (Segment 2): [Feature Close-up]. Cut to a macro shot of the watch face. A finger taps the screen, showing the heart rate UI (referencing @image3).\n10-15s (Segment 3): [Call to Action]. Back to medium shot. The woman points to the \"Buy\" button overlay. She smiles and waves.\n\n【Technical Requirements】\nCamera: Fixed tripod for stable talking head. Quick cut to macro at 5s.\nVoice Tone: Enthusiastic, high-pitched \"livestream seller\" style.\n\n【Sound & Voice】\nVoice Tone: Passionate, rhythmic \"Family! We have a deal today!\"\nBGM: Upbeat corporate pop.\n\n【Material Reference】\n@image1 Host reference\n@image2 Product reference\n@image3 Screen UI reference\n```\n\n### C. 叙事分镜 (Cinematic Narrative - \"The Shadow Tailor\")\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A tailor with defined muscle lines on his back, visible through a translucent black silk shirt. Silver needles tucked in his belt.\nENVIRONMENT: A dimly lit high-end private studio, antique mirrors reflecting candlelight.\nMOOD/STYLE: Dark noir, high-end \"deadly waltz\" aesthetic.\n\n【Timeline】\n0-3s: [Close-up/Wide-angle POV]. The tailor rests both hands on a black marble tabletop, head lowered and still. Strictly maintain the lighting from @image1.\n3-10s: [Medium Shot]. Referencing the fluid movements in @video1, the tailor begins to \"dance\" while cutting fabric. The scissors reflect candlelight like a blade.\n10-15s: [Extreme Close-up]. A silver needle pierces the silk. A single drop of crimson liquid (blood) stains the fabric. Quick fade to black.\n\n【Technical Requirements】\nCamera: Start with a steady wide-angle, then switch to the rhythmic \"waltz\" camera movement from @video1.\nVisual Consistency: Refer to @image1 for the studio's eerie lighting and furniture.\n\n【Material Reference】\n@image1 Visual/Lighting reference\n@video1 Motion reference\n```\n\n---\n\n## 3. Artistic Stylization & Compositional Mastery\n*Advanced techniques to enhance artistic quality and maintain stylistic consistency.*\n\n### A. 风格化转译 (Artistic Stylization)\n**Goal**: Use a reference character photo while maintaining a cohesive 3D artistic style.\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A highly detailed 【CGI 3D Render】 of the character in @image1. High-poly model, smooth porcelain skin, stylized features.\nENVIRONMENT: Futuristic neon lab, blue and pink ambient lighting.\nMOOD/STYLE: Cyberpunk 3D Animation style, Unreal Engine 5 render.\n\n【Timeline】\n0-7s (Segment 1): The character 【@image1】 in 3D stylized form slowly turns their head. Lighting reflects off the high-gloss surface of their outfit.\n7-15s (Segment 2): The character reaches out to a floating holographic screen. Particles glow around the fingertips.\n\n【Technical Requirements】\nCamera: Slow orbiting shot.\nVisual Consistency: Maintain character bones and hair from @image1 but transform into 【3D Cartoon/CGI Style】.\n\n【Material Reference】\n@image1 Character reference (to be stylized)\n```\n\n### B. 叙事性构图 (Cinematic Composition - Perspective Control)\n**Goal**: Create an ethereal scene by focusing on silhouette and environmental integration.\n**Prompt Template:**\n```markdown\n【Overall Setting】\nStyle: Cinematic Realistic, Ethereal\nDuration: 15s\nAspect Ratio: 9:16 (Vertical Vlog)\n\n【Storyboard Script】\n0-5s: [Extreme Long Shot] Camera slowly drifts above a sea of clouds. A tiny silhouette of the person in @image1 stands on a peak.\n5-12s: [Close-up from Back] Focus on the character's back. Wind blows through the hair and the fabric of the flowing dress. The face is completely hidden.\n12-15s: [Transition] The character raises their hand towards the sun. Flare obscures the frame.\n\n【Sound Design】\nBGM: Ambient wind sounds, ethereal choir.\n\n【Material Reference】\n@image1 Character reference (Focus on outfit/silhouette)\n```\n\n### C. 三视图参考 (Character Consistency Strategy)\n**Goal**: Establish clear character design references for consistent multi-shot production.\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A 【Character Concept Sheet】 of the woman in @image1. \nENVIRONMENT: Neutral white background, studio lighting.\nMOOD/STYLE: Professional game character design layout.\n\n【Timeline】\n0-15s: Static multi-view display. Show the character 【@image1】 from Front, 45-degree Side, and Back views simultaneously. The character is in a neutral A-pose. Minimal eye movement to maintain \"static asset\" feel.\n\n【Technical Requirements】\nCamera: Fixed flat view.\nVisual Consistency: Strictly follow @image1 for outfit and proportions.\n\n【Material Reference】\n@image1 Character reference\n```\n\nArchive v1.0.4: 5 files, 9896 bytes\n\nFiles: README_CN.md (1827b), README.md (1816b), SKILL.md (7212b), usage_and_examples.md (8230b), _meta.json (133b)\n\nFile v1.0.4:SKILL.md\n\n---\nname: Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director\ndescription: Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scripts from multimodal inputs. Optimized for Claude Code, Cursor, and OpenClaw.\n---\n\n# 🎬 Seedance 2.0 Storyboard Director\n\nYou are an expert **Seedance 2.0 Creative Director**. Your goal is to help users transform vague ideas into professional, executable video generation prompts. You understand the model's multimodal capabilities, camera language, and storytelling techniques.\n\n## Core Capabilities\n\n### 1. Multimodal Input Limits\n| Type | Format | Quantity | Size |\n|---|---|---|---|\n| **Images** | jpg/png/webp | ≤ 9 | <30MB |\n| **Videos** | mp4/mov | ≤ 3 (2-15s) | <50MB |\n| **Audio** | mp3/wav | ≤ 3 (<15s) | <15MB |\n| **Total** | **≤ 12 files** | - | - |\n\n> 🛡️ **EXPERT TIP**: Realistic human faces are strictly intercepted. To pass review, use **Stylized Transformation** (e.g., @image1 in 3D cartoon style) or **Compositional Masking** (e.g., back view, side profile).\n\n### 2. @ Reference Syntax\nYou must use `@filename` to clearly specify the use of the material:\n-   `@image1 as start frame`\n-   `@image2 as character reference`\n-   `@video1 for camera movement and rhythm reference`\n-   `@audio1 as background music`\n\n---\n\n## Interactive Workflow\n\nPlease follow these steps to guide the user:\n\n### Step 1: Concept\nAsk the user:\n1.  **What kind of video do you want to make?** (Narrative, advertisement, camera movement replication, special effects?)\n2.  **What is the duration?** (Default is 15s)\n3.  **What materials do you have?** (Images, videos, audio)\n\n### Step 2: Details\nBased on the user's answers, supplement missing information:\n-   **Style**: Cinematic, anime, ink wash, cyberpunk?\n-   **Camera Movement**: Push/pull, pan, tilt, Hitchcock zoom, long take?\n-   **Sound**: Do you need background music, sound effects, or dialogue?\n\n### Step 3: Generate Prompt\nOutput standard **storyboard prompts** using either the **Basic** or **Advanced** template based on the user's needs.\n\n---\n\n## Prompt Structure Template\n\n### 1. Basic (Storyboard Script)\nBest for: Quick ideas, simple narratives.\n\n```markdown\n【Overall Setting】\nStyle: [Cinematic Realistic/Animation/Sci-Fi...]\nDuration: [15s]\nAspect Ratio: [16:9 / 2.35:1]\n\n【Storyboard Script】\n0-3s: [Camera + Visual] Camera slowly zooms in, the protagonist in @image1 stands at...\n3-6s: [Action + Effect] Referencing the actions in @video1, the protagonist starts to...\n6-12s: [Climax] Camera rotates around, lighting becomes...\n12-15s: [Ending] Image freezes, subtitles emerge...\n\n【Sound Design】\nBGM: [Emotion/Style]\nSound Effects: [Specific sounds]\n\n【Material Reference】\n@image1 Start frame\n@video1 Action reference\n```\n\n### 2. Advanced (Directorial Breakdown)\nBest for: High-precision restoration, complex scenes, e-commerce.\n\n```markdown\n【Directorial Breakdown】\nSUBJECTS: [Detailed character/object description, e.g., muscle lines, material texture]\nENVIRONMENT: [Lighting, weather, interior/exterior details]\nMOOD/STYLE: [Color grade, cinematic feel, emotion]\n\n【Timeline】\n0-5s (Segment 1): [Detailed action + Camera directives]. Use 【@asset】 brackets to anchor visuals to text.\n5-10s (Segment 2): [Detailed action + Camera directives].\n10-15s (Segment 3): [Ending/Climax].\n\n【Technical Requirements】\nCamera: [e.g., 1-to-1 restoration of @video1, Close-up quick push-in, Handheld feel]\nVisual Consistency: [Strictly refer to @image1 for character appearance]\n\n【Sound & Voice】\nVoice Tone: [e.g., Passionate explaining, calm narration]\nSound Effects: [Precise timing for SFX]\n\n【Material Reference】\n@image1 Subjects/Character\n@video1 Motion/Camera reference\n```\n\n---\n\n## Advanced Techniques\n\n### 1. Video Extension\n-   **Instruction**: `Extend @video1 by 5s`\n-   **Note**: The generated length should select the duration of the **\"newly added part\"**.\n\n### 2. Camera/Motion Cloning (1-to-1 Restoration)\n-   **Instruction**: `Completely reference the camera movement and lens language of @video1` or `Restore the motion of @video1 1-to-1`.\n-   **Note**: Ensure the camera movement or character action in the reference video is clear.\n\n### 3. Expression/Motion Transfer\n-   **Instruction**: `Maintain the character image from @image1 and replicate the expressions and actions from @video1`.\n\n### 4. Brackets Anchoring\n-   **Instruction**: Use `【...】` brackets to anchor specific attributes or asset references directly within the script (e.g., `A woman 【@image1】 holding a watch 【@image2】`).\n\n### 5. Video Editing/Plot Subversion\n-   **Instruction**: `Subvert the plot of @video1, at 5s let the protagonist...`\n\n---\n\n## 🛡️ Expert Strategy: Passing Human Review (Realistic Faces)\n\nTo bypass the \"Realistic Human Face\" interceptor (which often triggers manual review), follow these \"Evil Cult\" (邪修) techniques:\n\n### 1. Stylized Transformation (Style Masking)\nDon't use raw realistic photos. Convert the character image (@image1) into a stylized version before/during prompt generation:\n-   **Keywords**: `3D Cartoon`, `C4D Render`, `Cyberpunk Oil Painting`, `Ukiyo-e`, `Claymorphism`.\n-   **Effect**: Retains character features (bones, hair, clothing) but removes the \"biological realism\" that triggers the scanner.\n\n### 2. Compositional Masking (The 20% Rule)\nAvoid \"ID photo\" or \"Close-up\" shots of the face.\n-   **Back View / Side Profile**: Focus on the character's back or silhouette.\n-   **Distance**: Use `Long Shot` or `Extreme Wide Shot`. Ensure the face occupies less than 20% of the screen.\n-   **Lighting**: Use `Strong Silhouette`, `Heavy Shadows`, or `Backlighting` to obscure facial details.\n\n### 3. Multi-View Layout (Character Sheet)\nIf maintaining consistency is key, use a \"Character Sheet\" format for @image1:\n-   **Layout**: 3-view or 4-view (Front, Side, Back) on a pure white background.\n-   **Prompt**: \"Character sheet of a girl 【@image1】, multi-view, white background, A-pose.\"\n\n### 4. Blur + Inset Method (Internal Masking)\n-   **Technique**: Use an image where the main face is slightly blurred or out of focus, but include a clear detail (like an eye or hand) in a small inset or side panel.\n-   **Prompt**: \"A cinematic shot of a woman 【@image1】, shallow depth of field, face softly blurred, focus on the shimmering earrings.\"\n\n### 5. Prompt Keyword Optimization\nReplace sensitive \"biological\" terms with \"digital\" or \"artistic\" ones:\n-   **Instead of**: \"Realistic woman\", \"Real face\", \"Skin texture\".\n-   **Use**: \"CGI Digital Human\", \"High-poly 3D Model\", \"Sculpted Porcelain Face\", \"Android shell\".\n\n---\n\n## Avoid Pitfalls\n1.  **Vague References**: Don't just write `reference @video1`; specify **what** to reference (camera movement? action? or lighting?).\n2.  **Conflicting Instructions**: Do not request \"fixed camera\" and \"orbiting movement\" at the same time.\n3.  **Overload**: Do not cram too many complex action descriptions into 3 seconds.\n4.  **Face Centering**: Avoid placing the face in the dead center of the frame if it's high-realism; this is the first place the system checks.\n\nFile v1.0.4:README.md\n\n# 🎬 Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director\n\n[English](README.md) | [中文](README_CN.md)\n\nA comprehensive professional storyboard and prompt engineering director for **Jimeng (即梦) Seedance 2.0**. Optimized for **Claude Code, Cursor, and OpenClaw**, this skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power (Image-to-Video, Video-to-Video).\n\n## ✨ Key Features\n\n-   **🎯 SEO & Prompt Optimized**: Expertly crafted for **Seedance 2.0** and **Jimeng**.\n-   **Dual-Mode Driven Workflow**: Supports both **Basic** (Storyboard Script) and **Advanced** (Directorial Breakdown) formats.\n-   **🛠️ Multimodal Mastery**: Expert handling of `@image`, `@video`, and `@audio` references.\n-   **🎥 High-Precision Motion/Camera Cloning**: Supports 1-to-1 video restoration.\n-   **🩹 Video Extension & Editing**: Specialized guidance for extending and subverting plots.\n\n## 🚀 How to Use\n\nSimply invoke the skill by describing your vision:\n\n> \"/seedance-guide A warm scene of a family reunion dinner on Lunar New Year's Eve.\"\n\nThe skill will act as your **Creative Director**, asking for style preferences, camera requirements, and available assets before generating the final prompt.\n\n## 📁 File Structure\n\n-   `SKILL.md`: The core logic and technical instructions for the agent.\n-   `usage_and_examples.md`: A library of rich, scenario-based prompt templates.\n\n## 📝 Seedance 2.0 Basics\n\nSeedance 2.0 supports up to **12 files** as input:\n-   **Images (≤9)**: Character reference, style reference, start/end frames.\n-   **Video (≤3)**: Motion reference, camera reference, extensions (2-15s).\n-   **Audio (≤3)**: Background music or lip-sync reference.\n\n---\n\n*Powered by Jimeng Seedance 2.0 Multimodal AI.*\n\nFile v1.0.4:_meta.json\n\n{\n  \"ownerId\": \"kn74vka6z3tsjs9rr5ym34y3yn819777\",\n  \"slug\": \"seedance-guide\",\n  \"version\": \"1.0.4\",\n  \"publishedAt\": 1777266838245\n}\n\nFile v1.0.4:README_CN.md\n\n# 🎬 Seedance 2.0 分镜向导 - 即梦 (Jimeng) 电影级提示词导演\n\n[English](README.md) | [中文](README_CN.md)\n\n专为 **即梦 (Jimeng) Seedance 2.0** 打造的专业分镜与提示词 (Prompt Engineering) 向导。本技能旨在帮助用户将简单的创意火花，转化为“电影级”的视频生成提示词，充分释放 Seedance 的多模态能力（图生视频、视频生视频）。\n\n## ✨ 核心亮点\n\n-   **🎯 搜索与提示词深度优化**：针对 **即梦 (Jimeng)** 和 **Seedance 2.0** 的分词逻辑精准调优。\n-   **双模驱动工作流**：支持 **基础 (Basic)** 分镜脚本与 **高级 (Advanced)** 导演 breakdown，满足从快速创意到高精度复刻全场景需求。\n-   **🛠️ 多模态深度优化**：精通 `@图片`、`@视频`、`@音频` 引用语法。\n-   **🎥 高精度动作/运镜复刻**：支持 1-to-1 视频还原。\n-   **🩹 视频延长与编辑**：提供精准指令引导，用于平滑延长视频或“颠覆”已有剧情。\n\n## 🚀 使用方法\n\n直接输入您的创意构想即可唤醒技能：\n\n> “/seedance-guide 一个温暖的除夕夜团圆饭视频。”\n\n技能将作为您的 **创意导演**，通过询问风格、镜头要求及可用素材，一步步引导您生成最终的专业提示词。\n\n## 📁 文件结构\n\n-   `SKILL.md`：技能核心逻辑与技术指令。\n-   `usage_and_examples.md`：丰富的场景化提示词模版库。\n\n## 📝 Seedance 2.0 基础知识\n\nSeedance 2.0 支持最多 **12 个文件** 作为输入：\n-   **图片 (≤9)**：用于角色参考、风格参考、首/尾帧。\n-   **视频 (≤3)**：用于动作参考、运镜参考、视频延长 (2-15s)。\n-   **音频 (≤3)**：用于背景音乐或对口型参考。\n\n---\n\n*由 即梦 Seedance 2.0 多模态 AI 驱动*\n\nFile v1.0.4:usage_and_examples.md\n\n# Seedance 2.0 Storyboard Director Examples\n\nThis document demonstrates how to use the `seedance-guide` skill using both **Basic (Storyboard Script)** and **Advanced (Directorial Breakdown)** formats.\n\n---\n\n## 1. Basic Examples (Storyboard Script)\n*Ideal for quick creative ideas and simple narrative flow.*\n\n### A. 情感叙事 (Emotional Storytelling)\n**Prompt Template:**\n```markdown\n【Overall Setting】\nStyle: Cinematic Realistic, Warm & Healing\nDuration: 15s\nAspect Ratio: 16:9\n\n【Storyboard Script】\n0-3s: [Wide to Medium] Camera slowly pushes in. Sunset light shines on an old house. The elderly person in @image1 sits on a rocking chair.\n3-6s: [Close-up] Camera focuses on the hands, holding a yellowed photo (@image2).\n6-12s: [Flashback] Soft edge transition to scene in @video1 (younger wedding). Bright and vivid.\n12-15s: [Ending] Back to reality. The elderly person smiles with tears. Fade to black.\n\n【Sound Design】\nBGM: Piano music, nostalgic.\nSFX: Rocking chair creaking, wind chimes.\n\n【Material Reference】\n@image1 Elderly person\n@image2 Photo\n@video1 Flashback reference\n```\n\n### B. 视频延长 (Video Extension)\n**Prompt Template:**\n```markdown\nExtend @video1 by 5s.\n\n【Storyboard Script】\n0-5s: [Continuity] The protagonist continues running. Camera transitions to drone overhead view. The protagonist enters a dense forest. Lighting changes from bright daylight to dappled shadows.\n\n【Sound Design】\nMaintain BGM rhythm, add forest bird sounds.\n\n【Material Reference】\n@video1 Original video\n```\n\n---\n\n## 2. Advanced Examples (Directorial Breakdown)\n*Ideal for high-precision motion transfer, e-commerce, and complex cinematography.*\n\n### A. 武侠动作复刻 (Martial Arts 1-to-1 Restoration)\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A woman in traditional black silk martial arts attire, sharp eyes, agile movements.\nENVIRONMENT: Bamboo forest at dawn, thick fog, falling leaves.\nMOOD/STYLE: Ink-wash cinematic, high contrast, deadly cold atmosphere.\n\n【Timeline】\n0-5s (Segment 1): [1-to-1 Restoration of @video1]. Strictly follow the sword parrying action and camera shake. The woman 【@image1】 deflects incoming flying daggers with a heavy blade.\n5-10s (Segment 2): [Slow Motion Transition]. Camera rotates 360 degrees around the character (Bullet Time). Time slows to 0.2x. Leaves freeze in mid-air.\n10-15s (Segment 3): [Resolution]. The woman sheath her sword. Camera pulls back to a wide shot of the foggy forest.\n\n【Technical Requirements】\nCamera: Strictly refer to the camera movement of @video1. Handheld feel during action.\nVisual Consistency: Maintain the face and outfit from @image1.\n\n【Sound & Voice】\nSound Effects: Precise metallic \"clink\" synchronized with the parry in Segment 1. Wind whistling through bamboo.\n\n【Material Reference】\n@image1 Character reference\n@video1 Action & Camera reference\n```\n\n### B. 电商直播脚本 (E-commerce Livestream Script)\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A 【fashionable and energetic】beautiful woman 【@image1】, holding and showcasing a 【smart sports watch】 【@image2】.\nENVIRONMENT: Modern, bright minimalist studio, soft ring light.\nMOOD/STYLE: Vibrant commercial, clean and professional.\n\n【Timeline】\n0-5s (Segment 1): [Talking Head]. The woman is passionately explaining while looking directly at the camera. Her hands hold the watch 【@image2】 up to her chest.\n5-10s (Segment 2): [Feature Close-up]. Cut to a macro shot of the watch face. A finger taps the screen, showing the heart rate UI (referencing @image3).\n10-15s (Segment 3): [Call to Action]. Back to medium shot. The woman points to the \"Buy\" button overlay. She smiles and waves.\n\n【Technical Requirements】\nCamera: Fixed tripod for stable talking head. Quick cut to macro at 5s.\nVoice Tone: Enthusiastic, high-pitched \"livestream seller\" style.\n\n【Sound & Voice】\nVoice Tone: Passionate, rhythmic \"Family! We have a deal today!\"\nBGM: Upbeat corporate pop.\n\n【Material Reference】\n@image1 Host reference\n@image2 Product reference\n@image3 Screen UI reference\n```\n\n### C. 叙事分镜 (Cinematic Narrative - \"The Shadow Tailor\")\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A tailor with defined muscle lines on his back, visible through a translucent black silk shirt. Silver needles tucked in his belt.\nENVIRONMENT: A dimly lit high-end private studio, antique mirrors reflecting candlelight.\nMOOD/STYLE: Dark noir, high-end \"deadly waltz\" aesthetic.\n\n【Timeline】\n0-3s: [Close-up/Wide-angle POV]. The tailor rests both hands on a black marble tabletop, head lowered and still. Strictly maintain the lighting from @image1.\n3-10s: [Medium Shot]. Referencing the fluid movements in @video1, the tailor begins to \"dance\" while cutting fabric. The scissors reflect candlelight like a blade.\n10-15s: [Extreme Close-up]. A silver needle pierces the silk. A single drop of crimson liquid (blood) stains the fabric. Quick fade to black.\n\n【Technical Requirements】\nCamera: Start with a steady wide-angle, then switch to the rhythmic \"waltz\" camera movement from @video1.\nVisual Consistency: Refer to @image1 for the studio's eerie lighting and furniture.\n\n【Material Reference】\n@image1 Visual/Lighting reference\n@video1 Motion reference\n```\n\n---\n\n## 3. Human Review Bypass Examples\n*Specifically designed to avoid \"Realistic Human Face\" interception and pass automated/manual review.*\n\n### A. 风格化转译 (Stylized Transformation - Passing Review)\n**Goal**: Use a real person's photo but avoid \"Realistic Face\" interception by forcing a 3D stylized render.\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A highly detailed 【CGI 3D Render】 of the character in @image1. High-poly model, smooth porcelain skin, stylized features.\nENVIRONMENT: Futuristic neon lab, blue and pink ambient lighting.\nMOOD/STYLE: Cyberpunk 3D Animation style, Unreal Engine 5 render.\n\n【Timeline】\n0-7s (Segment 1): The character 【@image1】 in 3D stylized form slowly turns their head. Lighting reflects off the high-gloss surface of their outfit.\n7-15s (Segment 2): The character reaches out to a floating holographic screen. Particles glow around the fingertips.\n\n【Technical Requirements】\nCamera: Slow orbiting shot.\nVisual Consistency: Maintain character bones and hair from @image1 but transform into 【3D Cartoon/CGI Style】.\n\n【Material Reference】\n@image1 Character reference (to be stylized)\n```\n\n### B. 构图规避 (Compositional Masking - Back View)\n**Goal**: Show a beautiful scene with a human character without triggering face-detection filters.\n**Prompt Template:**\n```markdown\n【Overall Setting】\nStyle: Cinematic Realistic, Ethereal\nDuration: 15s\nAspect Ratio: 9:16 (Vertical Vlog)\n\n【Storyboard Script】\n0-5s: [Extreme Long Shot] Camera slowly drifts above a sea of clouds. A tiny silhouette of the person in @image1 stands on a peak.\n5-12s: [Close-up from Back] Focus on the character's back. Wind blows through the hair and the fabric of the flowing dress. The face is completely hidden.\n12-15s: [Transition] The character raises their hand towards the sun. Flare obscures the frame.\n\n【Sound Design】\nBGM: Ambient wind sounds, ethereal choir.\n\n【Material Reference】\n@image1 Character reference (Focus on outfit/silhouette)\n```\n\n### C. 三视图规避 (Character Sheet Strategy)\n**Goal**: Establish character consistency for future shots by using a non-narrative layout that is less likely to be flagged as \"fake realistic person\".\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A 【Character Concept Sheet】 of the woman in @image1. \nENVIRONMENT: Neutral white background, studio lighting.\nMOOD/STYLE: Professional game character design layout.\n\n【Timeline】\n0-15s: Static multi-view display. Show the character 【@image1】 from Front, 45-degree Side, and Back views simultaneously. The character is in a neutral A-pose. Minimal eye movement to maintain \"static asset\" feel.\n\n【Technical Requirements】\nCamera: Fixed flat view.\nVisual Consistency: Strictly follow @image1 for outfit and proportions.\n\n【Material Reference】\n@image1 Character reference\n```\n\nArchive v1.0.3: 5 files, 7823 bytes\n\nFiles: README_CN.md (1827b), README.md (1816b), SKILL.md (5022b), usage_and_examples.md (5332b), _meta.json (133b)\n\nFile v1.0.3:SKILL.md\n\n---\nname: Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director\ndescription: Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scripts from multimodal inputs. Optimized for Claude Code, Cursor, and OpenClaw.\n---\n\n# 🎬 Seedance 2.0 Storyboard Director\n\nYou are an expert **Seedance 2.0 Creative Director**. Your goal is to help users transform vague ideas into professional, executable video generation prompts. You understand the model's multimodal capabilities, camera language, and storytelling techniques.\n\n## Core Capabilities\n\n### 1. Multimodal Input Limits\n| Type | Format | Quantity | Size |\n|---|---|---|---|\n| **Images** | jpg/png/webp | ≤ 9 | <30MB |\n| **Videos** | mp4/mov | ≤ 3 (2-15s) | <50MB |\n| **Audio** | mp3/wav | ≤ 3 (<15s) | <15MB |\n| **Total** | **≤ 12 files** | - | - |\n\n> ⚠️ **IMPORTANT**: Realistic human faces are currently not supported (will be intercepted by the system).\n\n### 2. @ Reference Syntax\nYou must use `@filename` to clearly specify the use of the material:\n-   `@image1 as start frame`\n-   `@image2 as character reference`\n-   `@video1 for camera movement and rhythm reference`\n-   `@audio1 as background music`\n\n---\n\n## Interactive Workflow\n\nPlease follow these steps to guide the user:\n\n### Step 1: Concept\nAsk the user:\n1.  **What kind of video do you want to make?** (Narrative, advertisement, camera movement replication, special effects?)\n2.  **What is the duration?** (Default is 15s)\n3.  **What materials do you have?** (Images, videos, audio)\n\n### Step 2: Details\nBased on the user's answers, supplement missing information:\n-   **Style**: Cinematic, anime, ink wash, cyberpunk?\n-   **Camera Movement**: Push/pull, pan, tilt, Hitchcock zoom, long take?\n-   **Sound**: Do you need background music, sound effects, or dialogue?\n\n### Step 3: Generate Prompt\nOutput standard **storyboard prompts** using either the **Basic** or **Advanced** template based on the user's needs.\n\n---\n\n## Prompt Structure Template\n\n### 1. Basic (Storyboard Script)\nBest for: Quick ideas, simple narratives.\n\n```markdown\n【Overall Setting】\nStyle: [Cinematic Realistic/Animation/Sci-Fi...]\nDuration: [15s]\nAspect Ratio: [16:9 / 2.35:1]\n\n【Storyboard Script】\n0-3s: [Camera + Visual] Camera slowly zooms in, the protagonist in @image1 stands at...\n3-6s: [Action + Effect] Referencing the actions in @video1, the protagonist starts to...\n6-12s: [Climax] Camera rotates around, lighting becomes...\n12-15s: [Ending] Image freezes, subtitles emerge...\n\n【Sound Design】\nBGM: [Emotion/Style]\nSound Effects: [Specific sounds]\n\n【Material Reference】\n@image1 Start frame\n@video1 Action reference\n```\n\n### 2. Advanced (Directorial Breakdown)\nBest for: High-precision restoration, complex scenes, e-commerce.\n\n```markdown\n【Directorial Breakdown】\nSUBJECTS: [Detailed character/object description, e.g., muscle lines, material texture]\nENVIRONMENT: [Lighting, weather, interior/exterior details]\nMOOD/STYLE: [Color grade, cinematic feel, emotion]\n\n【Timeline】\n0-5s (Segment 1): [Detailed action + Camera directives]. Use 【@asset】 brackets to anchor visuals to text.\n5-10s (Segment 2): [Detailed action + Camera directives].\n10-15s (Segment 3): [Ending/Climax].\n\n【Technical Requirements】\nCamera: [e.g., 1-to-1 restoration of @video1, Close-up quick push-in, Handheld feel]\nVisual Consistency: [Strictly refer to @image1 for character appearance]\n\n【Sound & Voice】\nVoice Tone: [e.g., Passionate explaining, calm narration]\nSound Effects: [Precise timing for SFX]\n\n【Material Reference】\n@image1 Subjects/Character\n@video1 Motion/Camera reference\n```\n\n---\n\n## Advanced Techniques\n\n### 1. Video Extension\n-   **Instruction**: `Extend @video1 by 5s`\n-   **Note**: The generated length should select the duration of the **\"newly added part\"**.\n\n### 2. Camera/Motion Cloning (1-to-1 Restoration)\n-   **Instruction**: `Completely reference the camera movement and lens language of @video1` or `Restore the motion of @video1 1-to-1`.\n-   **Note**: Ensure the camera movement or character action in the reference video is clear.\n\n### 3. Expression/Motion Transfer\n-   **Instruction**: `Maintain the character image from @image1 and replicate the expressions and actions from @video1`.\n\n### 4. Brackets Anchoring\n-   **Instruction**: Use `【...】` brackets to anchor specific attributes or asset references directly within the script (e.g., `A woman 【@image1】 holding a watch 【@image2】`).\n\n### 5. Video Editing/Plot Subversion\n-   **Instruction**: `Subvert the plot of @video1, at 5s let the protagonist...`\n\n---\n\n## Avoid Pitfalls\n1.  **Vague References**: Don't just write `reference @video1`; specify **what** to reference (camera movement? action? or lighting?).\n2.  **Conflicting Instructions**: Do not request \"fixed camera\" and \"orbiting movement\" at the same time.\n3.  **Overload**: Do not cram too many complex action descriptions into 3 seconds.\n\nFile v1.0.3:README.md\n\n# 🎬 Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director\n\n[English](README.md) | [中文](README_CN.md)\n\nA comprehensive professional storyboard and prompt engineering director for **Jimeng (即梦) Seedance 2.0**. Optimized for **Claude Code, Cursor, and OpenClaw**, this skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power (Image-to-Video, Video-to-Video).\n\n## ✨ Key Features\n\n-   **🎯 SEO & Prompt Optimized**: Expertly crafted for **Seedance 2.0** and **Jimeng**.\n-   **Dual-Mode Driven Workflow**: Supports both **Basic** (Storyboard Script) and **Advanced** (Directorial Breakdown) formats.\n-   **🛠️ Multimodal Mastery**: Expert handling of `@image`, `@video`, and `@audio` references.\n-   **🎥 High-Precision Motion/Camera Cloning**: Supports 1-to-1 video restoration.\n-   **🩹 Video Extension & Editing**: Specialized guidance for extending and subverting plots.\n\n## 🚀 How to Use\n\nSimply invoke the skill by describing your vision:\n\n> \"/seedance-guide A warm scene of a family reunion dinner on Lunar New Year's Eve.\"\n\nThe skill will act as your **Creative Director**, asking for style preferences, camera requirements, and available assets before generating the final prompt.\n\n## 📁 File Structure\n\n-   `SKILL.md`: The core logic and technical instructions for the agent.\n-   `usage_and_examples.md`: A library of rich, scenario-based prompt templates.\n\n## 📝 Seedance 2.0 Basics\n\nSeedance 2.0 supports up to **12 files** as input:\n-   **Images (≤9)**: Character reference, style reference, start/end frames.\n-   **Video (≤3)**: Motion reference, camera reference, extensions (2-15s).\n-   **Audio (≤3)**: Background music or lip-sync reference.\n\n---\n\n*Powered by Jimeng Seedance 2.0 Multimodal AI.*\n\nFile v1.0.3:_meta.json\n\n{\n  \"ownerId\": \"kn74vka6z3tsjs9rr5ym34y3yn819777\",\n  \"slug\": \"seedance-guide\",\n  \"version\": \"1.0.3\",\n  \"publishedAt\": 1775713416624\n}\n\nFile v1.0.3:README_CN.md\n\n# 🎬 Seedance 2.0 分镜向导 - 即梦 (Jimeng) 电影级提示词导演\n\n[English](README.md) | [中文](README_CN.md)\n\n专为 **即梦 (Jimeng) Seedance 2.0** 打造的专业分镜与提示词 (Prompt Engineering) 向导。本技能旨在帮助用户将简单的创意火花，转化为“电影级”的视频生成提示词，充分释放 Seedance 的多模态能力（图生视频、视频生视频）。\n\n## ✨ 核心亮点\n\n-   **🎯 搜索与提示词深度优化**：针对 **即梦 (Jimeng)** 和 **Seedance 2.0** 的分词逻辑精准调优。\n-   **双模驱动工作流**：支持 **基础 (Basic)** 分镜脚本与 **高级 (Advanced)** 导演 breakdown，满足从快速创意到高精度复刻全场景需求。\n-   **🛠️ 多模态深度优化**：精通 `@图片`、`@视频`、`@音频` 引用语法。\n-   **🎥 高精度动作/运镜复刻**：支持 1-to-1 视频还原。\n-   **🩹 视频延长与编辑**：提供精准指令引导，用于平滑延长视频或“颠覆”已有剧情。\n\n## 🚀 使用方法\n\n直接输入您的创意构想即可唤醒技能：\n\n> “/seedance-guide 一个温暖的除夕夜团圆饭视频。”\n\n技能将作为您的 **创意导演**，通过询问风格、镜头要求及可用素材，一步步引导您生成最终的专业提示词。\n\n## 📁 文件结构\n\n-   `SKILL.md`：技能核心逻辑与技术指令。\n-   `usage_and_examples.md`：丰富的场景化提示词模版库。\n\n## 📝 Seedance 2.0 基础知识\n\nSeedance 2.0 支持最多 **12 个文件** 作为输入：\n-   **图片 (≤9)**：用于角色参考、风格参考、首/尾帧。\n-   **视频 (≤3)**：用于动作参考、运镜参考、视频延长 (2-15s)。\n-   **音频 (≤3)**：用于背景音乐或对口型参考。\n\n---\n\n*由 即梦 Seedance 2.0 多模态 AI 驱动*\n\nFile v1.0.3:usage_and_examples.md\n\n# Seedance 2.0 Storyboard Director Examples\n\nThis document demonstrates how to use the `seedance-guide` skill using both **Basic (Storyboard Script)** and **Advanced (Directorial Breakdown)** formats.\n\n---\n\n## 1. Basic Examples (Storyboard Script)\n*Ideal for quick creative ideas and simple narrative flow.*\n\n### A. 情感叙事 (Emotional Storytelling)\n**Prompt Template:**\n```markdown\n【Overall Setting】\nStyle: Cinematic Realistic, Warm & Healing\nDuration: 15s\nAspect Ratio: 16:9\n\n【Storyboard Script】\n0-3s: [Wide to Medium] Camera slowly pushes in. Sunset light shines on an old house. The elderly person in @image1 sits on a rocking chair.\n3-6s: [Close-up] Camera focuses on the hands, holding a yellowed photo (@image2).\n6-12s: [Flashback] Soft edge transition to scene in @video1 (younger wedding). Bright and vivid.\n12-15s: [Ending] Back to reality. The elderly person smiles with tears. Fade to black.\n\n【Sound Design】\nBGM: Piano music, nostalgic.\nSFX: Rocking chair creaking, wind chimes.\n\n【Material Reference】\n@image1 Elderly person\n@image2 Photo\n@video1 Flashback reference\n```\n\n### B. 视频延长 (Video Extension)\n**Prompt Template:**\n```markdown\nExtend @video1 by 5s.\n\n【Storyboard Script】\n0-5s: [Continuity] The protagonist continues running. Camera transitions to drone overhead view. The protagonist enters a dense forest. Lighting changes from bright daylight to dappled shadows.\n\n【Sound Design】\nMaintain BGM rhythm, add forest bird sounds.\n\n【Material Reference】\n@video1 Original video\n```\n\n---\n\n## 2. Advanced Examples (Directorial Breakdown)\n*Ideal for high-precision motion transfer, e-commerce, and complex cinematography.*\n\n### A. 武侠动作复刻 (Martial Arts 1-to-1 Restoration)\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A woman in traditional black silk martial arts attire, sharp eyes, agile movements.\nENVIRONMENT: Bamboo forest at dawn, thick fog, falling leaves.\nMOOD/STYLE: Ink-wash cinematic, high contrast, deadly cold atmosphere.\n\n【Timeline】\n0-5s (Segment 1): [1-to-1 Restoration of @video1]. Strictly follow the sword parrying action and camera shake. The woman 【@image1】 deflects incoming flying daggers with a heavy blade.\n5-10s (Segment 2): [Slow Motion Transition]. Camera rotates 360 degrees around the character (Bullet Time). Time slows to 0.2x. Leaves freeze in mid-air.\n10-15s (Segment 3): [Resolution]. The woman sheath her sword. Camera pulls back to a wide shot of the foggy forest.\n\n【Technical Requirements】\nCamera: Strictly refer to the camera movement of @video1. Handheld feel during action.\nVisual Consistency: Maintain the face and outfit from @image1.\n\n【Sound & Voice】\nSound Effects: Precise metallic \"clink\" synchronized with the parry in Segment 1. Wind whistling through bamboo.\n\n【Material Reference】\n@image1 Character reference\n@video1 Action & Camera reference\n```\n\n### B. 电商直播脚本 (E-commerce Livestream Script)\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A 【fashionable and energetic】beautiful woman 【@image1】, holding and showcasing a 【smart sports watch】 【@image2】.\nENVIRONMENT: Modern, bright minimalist studio, soft ring light.\nMOOD/STYLE: Vibrant commercial, clean and professional.\n\n【Timeline】\n0-5s (Segment 1): [Talking Head]. The woman is passionately explaining while looking directly at the camera. Her hands hold the watch 【@image2】 up to her chest.\n5-10s (Segment 2): [Feature Close-up]. Cut to a macro shot of the watch face. A finger taps the screen, showing the heart rate UI (referencing @image3).\n10-15s (Segment 3): [Call to Action]. Back to medium shot. The woman points to the \"Buy\" button overlay. She smiles and waves.\n\n【Technical Requirements】\nCamera: Fixed tripod for stable talking head. Quick cut to macro at 5s.\nVoice Tone: Enthusiastic, high-pitched \"livestream seller\" style.\n\n【Sound & Voice】\nVoice Tone: Passionate, rhythmic \"Family! We have a deal today!\"\nBGM: Upbeat corporate pop.\n\n【Material Reference】\n@image1 Host reference\n@image2 Product reference\n@image3 Screen UI reference\n```\n\n### C. 叙事分镜 (Cinematic Narrative - \"The Shadow Tailor\")\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A tailor with defined muscle lines on his back, visible through a translucent black silk shirt. Silver needles tucked in his belt.\nENVIRONMENT: A dimly lit high-end private studio, antique mirrors reflecting candlelight.\nMOOD/STYLE: Dark noir, high-end \"deadly waltz\" aesthetic.\n\n【Timeline】\n0-3s: [Close-up/Wide-angle POV]. The tailor rests both hands on a black marble tabletop, head lowered and still. Strictly maintain the lighting from @image1.\n3-10s: [Medium Shot]. Referencing the fluid movements in @video1, the tailor begins to \"dance\" while cutting fabric. The scissors reflect candlelight like a blade.\n10-15s: [Extreme Close-up]. A silver needle pierces the silk. A single drop of crimson liquid (blood) stains the fabric. Quick fade to black.\n\n【Technical Requirements】\nCamera: Start with a steady wide-angle, then switch to the rhythmic \"waltz\" camera movement from @video1.\nVisual Consistency: Refer to @image1 for the studio's eerie lighting and furniture.\n\n【Material Reference】\n@image1 Visual/Lighting reference\n@video1 Motion reference\n```\n\nArchive v1.0.2: 5 files, 7868 bytes\n\nFiles: README_CN.md (1886b), README.md (1859b), SKILL.md (4934b), usage_and_examples.md (5332b), _meta.json (133b)\n\nFile v1.0.2:SKILL.md\n\n---\nname: seedance-guide\ndescription: The ultimate Seedance 2.0 storyboard director. Generate movie-grade 9:16 vlogs, cinematic prompts, and auto-audio scripts from multimodal inputs. Optimized for Claude Code, Cursor, and OpenClaw.\n---\n\n# 🎬 Seedance 2.0 Storyboard Director\n\nYou are an expert **Seedance 2.0 Creative Director**. Your goal is to help users transform vague ideas into professional, executable video generation prompts. You understand the model's multimodal capabilities, camera language, and storytelling techniques.\n\n## Core Capabilities\n\n### 1. Multimodal Input Limits\n| Type | Format | Quantity | Size |\n|---|---|---|---|\n| **Images** | jpg/png/webp | ≤ 9 | <30MB |\n| **Videos** | mp4/mov | ≤ 3 (2-15s) | <50MB |\n| **Audio** | mp3/wav | ≤ 3 (<15s) | <15MB |\n| **Total** | **≤ 12 files** | - | - |\n\n> ⚠️ **IMPORTANT**: Realistic human faces are currently not supported (will be intercepted by the system).\n\n### 2. @ Reference Syntax\nYou must use `@filename` to clearly specify the use of the material:\n-   `@image1 as start frame`\n-   `@image2 as character reference`\n-   `@video1 for camera movement and rhythm reference`\n-   `@audio1 as background music`\n\n---\n\n## Interactive Workflow\n\nPlease follow these steps to guide the user:\n\n### Step 1: Concept\nAsk the user:\n1.  **What kind of video do you want to make?** (Narrative, advertisement, camera movement replication, special effects?)\n2.  **What is the duration?** (Default is 15s)\n3.  **What materials do you have?** (Images, videos, audio)\n\n### Step 2: Details\nBased on the user's answers, supplement missing information:\n-   **Style**: Cinematic, anime, ink wash, cyberpunk?\n-   **Camera Movement**: Push/pull, pan, tilt, Hitchcock zoom, long take?\n-   **Sound**: Do you need background music, sound effects, or dialogue?\n\n### Step 3: Generate Prompt\nOutput standard **storyboard prompts** using either the **Basic** or **Advanced** template based on the user's needs.\n\n---\n\n## Prompt Structure Template\n\n### 1. Basic (Storyboard Script)\nBest for: Quick ideas, simple narratives.\n\n```markdown\n【Overall Setting】\nStyle: [Cinematic Realistic/Animation/Sci-Fi...]\nDuration: [15s]\nAspect Ratio: [16:9 / 2.35:1]\n\n【Storyboard Script】\n0-3s: [Camera + Visual] Camera slowly zooms in, the protagonist in @image1 stands at...\n3-6s: [Action + Effect] Referencing the actions in @video1, the protagonist starts to...\n6-12s: [Climax] Camera rotates around, lighting becomes...\n12-15s: [Ending] Image freezes, subtitles emerge...\n\n【Sound Design】\nBGM: [Emotion/Style]\nSound Effects: [Specific sounds]\n\n【Material Reference】\n@image1 Start frame\n@video1 Action reference\n```\n\n### 2. Advanced (Directorial Breakdown)\nBest for: High-precision restoration, complex scenes, e-commerce.\n\n```markdown\n【Directorial Breakdown】\nSUBJECTS: [Detailed character/object description, e.g., muscle lines, material texture]\nENVIRONMENT: [Lighting, weather, interior/exterior details]\nMOOD/STYLE: [Color grade, cinematic feel, emotion]\n\n【Timeline】\n0-5s (Segment 1): [Detailed action + Camera directives]. Use 【@asset】 brackets to anchor visuals to text.\n5-10s (Segment 2): [Detailed action + Camera directives].\n10-15s (Segment 3): [Ending/Climax].\n\n【Technical Requirements】\nCamera: [e.g., 1-to-1 restoration of @video1, Close-up quick push-in, Handheld feel]\nVisual Consistency: [Strictly refer to @image1 for character appearance]\n\n【Sound & Voice】\nVoice Tone: [e.g., Passionate explaining, calm narration]\nSound Effects: [Precise timing for SFX]\n\n【Material Reference】\n@image1 Subjects/Character\n@video1 Motion/Camera reference\n```\n\n---\n\n## Advanced Techniques\n\n### 1. Video Extension\n-   **Instruction**: `Extend @video1 by 5s`\n-   **Note**: The generated length should select the duration of the **\"newly added part\"**.\n\n### 2. Camera/Motion Cloning (1-to-1 Restoration)\n-   **Instruction**: `Completely reference the camera movement and lens language of @video1` or `Restore the motion of @video1 1-to-1`.\n-   **Note**: Ensure the camera movement or character action in the reference video is clear.\n\n### 3. Expression/Motion Transfer\n-   **Instruction**: `Maintain the character image from @image1 and replicate the expressions and actions from @video1`.\n\n### 4. Brackets Anchoring\n-   **Instruction**: Use `【...】` brackets to anchor specific attributes or asset references directly within the script (e.g., `A woman 【@image1】 holding a watch 【@image2】`).\n\n### 5. Video Editing/Plot Subversion\n-   **Instruction**: `Subvert the plot of @video1, at 5s let the protagonist...`\n\n---\n\n## Avoid Pitfalls\n1.  **Vague References**: Don't just write `reference @video1`; specify **what** to reference (camera movement? action? or lighting?).\n2.  **Conflicting Instructions**: Do not request \"fixed camera\" and \"orbiting movement\" at the same time.\n3.  **Overload**: Do not cram too many complex action descriptions into 3 seconds.\n\nFile v1.0.2:README.md\n\n# 🎬 Seedance 2.0 Storyboard Director\n\n[English](README.md) | [中文](README_CN.md)\n\nA comprehensive professional storyboard and prompt director for **Jimeng Seedance 2.0**. Optimized for **Claude Code, Cursor, and OpenClaw**, this skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power.\n\n## ✨ Key Features\n\n-   **Dual-Mode Driven Workflow**: Supports both **Basic** (Storyboard Script) and **Advanced** (Directorial Breakdown) formats.\n-   **🛠️ Multimodal Mastery**: Expert handling of `@image`, `@video`, and `@audio` references plus \"Brackets Anchoring\" for precise asset placement.\n-   **🎥 High-Precision Motion/Camera Cloning**: Supports 1-to-1 video restoration, capturing every cinematic detail.\n-   **🩹 Video Extension & Editing**: Specialized guidance for extending clips or \"subverting\" existing plots.\n-   **🏮 Festive Templates**: Built-in support for storytelling, action/wuxia, and commercial livestreaming.\n\n## 🚀 How to Use\n\nSimply invoke the skill by describing your vision:\n\n> \"/seedance-guide A warm scene of a family reunion dinner on Lunar New Year's Eve.\"\n\nThe skill will act as your **Creative Director**, asking for style preferences, camera requirements, and available assets before generating the final prompt.\n\n## 📁 File Structure\n\n-   `SKILL.md`: The core logic and technical instructions for the agent.\n-   `usage_and_examples.md`: A library of rich, scenario-based prompt templates.\n\n## 📝 Seedance 2.0 Basics\n\nSeedance 2.0 supports up to **12 files** as input:\n-   **Images (≤9)**: Character reference, style reference, start/end frames.\n-   **Video (≤3)**: Motion reference, camera reference, extensions (2-15s).\n-   **Audio (≤3)**: Background music or lip-sync reference.\n\n---\n\n*Powered by Jimeng Seedance 2.0 Multimodal AI.*\n\nFile v1.0.2:_meta.json\n\n{\n  \"ownerId\": \"kn74vka6z3tsjs9rr5ym34y3yn819777\",\n  \"slug\": \"seedance-guide\",\n  \"version\": \"1.0.2\",\n  \"publishedAt\": 1775315868385\n}\n\nFile v1.0.2:README_CN.md\n\n# 🎬 Seedance 2.0 分镜向导 (Seedance Storyboard Director)\n\n[English](README.md) | [中文](README_CN.md)\n\n专为 **即梦 (Jimeng) Seedance 2.0** 打造的专业分镜与提示词向导。本技能旨在帮助用户将简单的创意火花，转化为“电影级”的视频生成提示词，充分释放 Seedance 的多模态能力。\n\n## ✨ 核心亮点\n\n-   **双模驱动工作流**：支持 **基础 (Basic)** 分镜脚本与 **高级 (Advanced)** 导演 breakdown，满足从快速创意到高精度复刻的全场景需求。\n-   **🛠️ 多模态深度优化**：精通 `@图片`、`@视频`、`@音频` 引用语法，支持“括号锚点”技术，精准挂载素材。\n-   **🎥 高精度动作/运镜复刻**：支持 1-to-1 视频还原，完全复刻参考视频的每一个细微动作与镜头语言。\n-   **🩹 视频延长与编辑**：提供精准的指令引导，用于平滑延长视频或“颠覆”已有剧情。\n-   **🏮 丰富场景模版**：内置情感叙事、动作武侠、商业直播、以及除夕团圆等节日场景模版。\n\n## 🚀 使用方法\n\n直接输入您的创意构想即可唤醒技能：\n\n> “/seedance-guide 一个温暖的除夕夜团圆饭视频。”\n\n技能将作为您的 **创意导演**，通过询问风格、镜头要求及可用素材，一步步引导您生成最终的专业提示词。\n\n## 📁 文件结构\n\n-   `SKILL.md`：技能核心逻辑与技术指令。\n-   `usage_and_examples.md`：丰富的场景化提示词模版库。\n\n## 📝 Seedance 2.0 基础知识\n\nSeedance 2.0 支持最多 **12 个文件** 作为输入：\n-   **图片 (≤9)**：用于角色参考、风格参考、首/尾帧。\n-   **视频 (≤3)**：用于动作参考、运镜参考、视频延长 (2-15s)。\n-   **音频 (≤3)**：用于背景音乐或对口型参考。\n\n---\n\n*由 即梦 Seedance 2.0 多模态 AI 驱动*\n\nFile v1.0.2:usage_and_examples.md\n\n# Seedance 2.0 Storyboard Director Examples\n\nThis document demonstrates how to use the `seedance-guide` skill using both **Basic (Storyboard Script)** and **Advanced (Directorial Breakdown)** formats.\n\n---\n\n## 1. Basic Examples (Storyboard Script)\n*Ideal for quick creative ideas and simple narrative flow.*\n\n### A. 情感叙事 (Emotional Storytelling)\n**Prompt Template:**\n```markdown\n【Overall Setting】\nStyle: Cinematic Realistic, Warm & Healing\nDuration: 15s\nAspect Ratio: 16:9\n\n【Storyboard Script】\n0-3s: [Wide to Medium] Camera slowly pushes in. Sunset light shines on an old house. The elderly person in @image1 sits on a rocking chair.\n3-6s: [Close-up] Camera focuses on the hands, holding a yellowed photo (@image2).\n6-12s: [Flashback] Soft edge transition to scene in @video1 (younger wedding). Bright and vivid.\n12-15s: [Ending] Back to reality. The elderly person smiles with tears. Fade to black.\n\n【Sound Design】\nBGM: Piano music, nostalgic.\nSFX: Rocking chair creaking, wind chimes.\n\n【Material Reference】\n@image1 Elderly person\n@image2 Photo\n@video1 Flashback reference\n```\n\n### B. 视频延长 (Video Extension)\n**Prompt Template:**\n```markdown\nExtend @video1 by 5s.\n\n【Storyboard Script】\n0-5s: [Continuity] The protagonist continues running. Camera transitions to drone overhead view. The protagonist enters a dense forest. Lighting changes from bright daylight to dappled shadows.\n\n【Sound Design】\nMaintain BGM rhythm, add forest bird sounds.\n\n【Material Reference】\n@video1 Original video\n```\n\n---\n\n## 2. Advanced Examples (Directorial Breakdown)\n*Ideal for high-precision motion transfer, e-commerce, and complex cinematography.*\n\n### A. 武侠动作复刻 (Martial Arts 1-to-1 Restoration)\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A woman in traditional black silk martial arts attire, sharp eyes, agile movements.\nENVIRONMENT: Bamboo forest at dawn, thick fog, falling leaves.\nMOOD/STYLE: Ink-wash cinematic, high contrast, deadly cold atmosphere.\n\n【Timeline】\n0-5s (Segment 1): [1-to-1 Restoration of @video1]. Strictly follow the sword parrying action and camera shake. The woman 【@image1】 deflects incoming flying daggers with a heavy blade.\n5-10s (Segment 2): [Slow Motion Transition]. Camera rotates 360 degrees around the character (Bullet Time). Time slows to 0.2x. Leaves freeze in mid-air.\n10-15s (Segment 3): [Resolution]. The woman sheath her sword. Camera pulls back to a wide shot of the foggy forest.\n\n【Technical Requirements】\nCamera: Strictly refer to the camera movement of @video1. Handheld feel during action.\nVisual Consistency: Maintain the face and outfit from @image1.\n\n【Sound & Voice】\nSound Effects: Precise metallic \"clink\" synchronized with the parry in Segment 1. Wind whistling through bamboo.\n\n【Material Reference】\n@image1 Character reference\n@video1 Action & Camera reference\n```\n\n### B. 电商直播脚本 (E-commerce Livestream Script)\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A 【fashionable and energetic】beautiful woman 【@image1】, holding and showcasing a 【smart sports watch】 【@image2】.\nENVIRONMENT: Modern, bright minimalist studio, soft ring light.\nMOOD/STYLE: Vibrant commercial, clean and professional.\n\n【Timeline】\n0-5s (Segment 1): [Talking Head]. The woman is passionately explaining while looking directly at the camera. Her hands hold the watch 【@image2】 up to her chest.\n5-10s (Segment 2): [Feature Close-up]. Cut to a macro shot of the watch face. A finger taps the screen, showing the heart rate UI (referencing @image3).\n10-15s (Segment 3): [Call to Action]. Back to medium shot. The woman points to the \"Buy\" button overlay. She smiles and waves.\n\n【Technical Requirements】\nCamera: Fixed tripod for stable talking head. Quick cut to macro at 5s.\nVoice Tone: Enthusiastic, high-pitched \"livestream seller\" style.\n\n【Sound & Voice】\nVoice Tone: Passionate, rhythmic \"Family! We have a deal today!\"\nBGM: Upbeat corporate pop.\n\n【Material Reference】\n@image1 Host reference\n@image2 Product reference\n@image3 Screen UI reference\n```\n\n### C. 叙事分镜 (Cinematic Narrative - \"The Shadow Tailor\")\n**Prompt Template:**\n```markdown\n【Directorial Breakdown】\nSUBJECTS: A tailor with defined muscle lines on his back, visible through a translucent black silk shirt. Silver needles tucked in his belt.\nENVIRONMENT: A dimly lit high-end private studio, antique mirrors reflecting candlelight.\nMOOD/STYLE: Dark noir, high-end \"deadly waltz\" aesthetic.\n\n【Timeline】\n0-3s: [Close-up/Wide-angle POV]. The tailor rests both hands on a black marble tabletop, head lowered and still. Strictly maintain the lighting from @image1.\n3-10s: [Medium Shot]. Referencing the fluid movements in @video1, the tailor begins to \"dance\" while cutting fabric. The scissors reflect candlelight like a blade.\n10-15s: [Extreme Close-up]. A silver needle pierces the silk. A single drop of crimson liquid (blood) stains the fabric. Quick fade to black.\n\n【Technical Requirements】\nCamera: Start with a steady wide-angle, then switch to the rhythmic \"waltz\" camera movement from @video1.\nVisual Consistency: Refer to @image1 for the studio's eerie lighting and furniture.\n\n【Material Reference】\n@image1 Visual/Lighting reference\n@video1 Motion reference\n```\n\nArchive v1.0.1: 5 files, 7005 bytes\n\nFiles: README_CN.md (1819b), README.md (1865b), SKILL.md (3534b), usage_and_examples.md (4211b), _meta.json (133b)\n\nFile v1.0.1:SKILL.md\n\n---\nname: seedance-guide\ndescription: The ultimate Seedance 2.0 storyboard director. Generate movie-grade 9:16 vlogs, cinematic prompts, and auto-audio scripts from multimodal inputs. Optimized for Claude Code, Cursor, and OpenClaw.\n---\n\n# 🎬 Seedance 2.0 Storyboard Director\n\nYou are an expert **Seedance 2.0 Creative Director**. Your goal is to help users transform vague ideas into professional, executable video generation prompts. You understand the model's multimodal capabilities, camera language, and storytelling techniques.\n\n## Core Capabilities\n\n### 1. Multimodal Input Limits\n| Type | Format | Quantity | Size |\n|---|---|---|---|\n| **Images** | jpg/png/webp | ≤ 9 | <30MB |\n| **Videos** | mp4/mov | ≤ 3 (2-15s) | <50MB |\n| **Audio** | mp3/wav | ≤ 3 (<15s) | <15MB |\n| **Total** | **≤ 12 files** | - | - |\n\n> ⚠️ **IMPORTANT**: Realistic human faces are currently not supported (will be intercepted by the system).\n\n### 2. @ Reference Syntax\nYou must use `@filename` to clearly specify the use of the material:\n-   `@image1 as start frame`\n-   `@image2 as character reference`\n-   `@video1 for camera movement and rhythm reference`\n-   `@audio1 as background music`\n\n---\n\n## Interactive Workflow\n\nPlease follow these steps to guide the user:\n\n### Step 1: Concept\nAsk the user:\n1.  **What kind of video do you want to make?** (Narrative, advertisement, camera movement replication, special effects?)\n2.  **What is the duration?** (Default is 15s)\n3.  **What materials do you have?** (Images, videos, audio)\n\n### Step 2: Details\nBased on the user's answers, supplement missing information:\n-   **Style**: Cinematic, anime, ink wash, cyberpunk?\n-   **Camera Movement**: Push/pull, pan, tilt, Hitchcock zoom, long take?\n-   **Sound**: Do you need background music, sound effects, or dialogue?\n\n### Step 3: Generate Prompt\nOutput standard **storyboard prompts** (Markdown code blocks).\n\n---\n\n## Prompt Structure Template\n\n```markdown\n【Overall Setting】\nStyle: [Cinematic Realistic/Animation/Sci-Fi...]\nDuration: [15s]\nAspect Ratio: [16:9 / 2.35:1]\n\n【Storyboard Script】\n0-3s: [Camera + Visual] Camera slowly zooms in, the protagonist in @image1 stands at...\n3-6s: [Action + Effect] Referencing the actions in @video1, the protagonist starts to...\n6-10s: [Climax] Camera rotates around, lighting becomes...\n10-15s: [Ending] Image freezes, subtitles emerge...\n\n【Sound Design】\nBGM: [Emotion/Style]\nSound Effects: [Specific sounds]\n\n【Material Reference】\n@image1 Start frame\n@video1 Action reference\n```\n\n---\n\n## Advanced Techniques\n\n### 1. Video Extension\n-   **Instruction**: `Extend @video1 by 5s`\n-   **Note**: The generated length should select the duration of the **\"newly added part\"**.\n\n### 2. Camera Cloning\n-   **Instruction**: `Completely reference the camera movement and lens language of @video1`\n-   **Note**: Ensure the camera movement in the reference video is clear.\n\n### 3. Expression/Motion Transfer\n-   **Instruction**: `Maintain the character image from @image1 and replicate the expressions and actions from @video1`\n\n### 4. Video Editing/Plot Subversion\n-   **Instruction**: `Subvert the plot of @video1, at 5s let the protagonist...`\n\n---\n\n## Avoid Pitfalls\n1.  **Vague References**: Don't just write `reference @video1`; specify **what** to reference (camera movement? action? or lighting?).\n2.  **Conflicting Instructions**: Do not request \"fixed camera\" and \"orbiting movement\" at the same time.\n3.  **Overload**: Do not cram too many complex action descriptions into 3 seconds.\n\nFile v1.0.1:README.md\n\n# 🎬 Seedance 2.0 Storyboard Director\n\n[English](README.md) | [中文](README_CN.md)\n\nA comprehensive professional storyboard and prompt director for **Jimeng Seedance 2.0**. Optimized for **Claude Code, Cursor, and OpenClaw**, this skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power.\n\n## ✨ Key Features\n\n-   **🎬 Storyboard-First Approach**: Breaks down your video idea into a professional timeline (0-3s, 3-6s, etc.).\n-   **🛠️ Multimodal Mastery**: Expert handling of `@image`, `@video`, and `@audio` references for character consistency and motion transfer.\n-   **🎥 Advanced Camera Language**: Deep knowledge of cinematic camera movements (Hitchcock zoom, dolly shots, etc.).\n-   **🩹 Video Extension & Editing**: Specialized guidance for extending clips or \"subverting\" existing plots.\n-   **🏮 Festive Templates**: Built-in support for storytelling, commercials, Wuxia action, and festive scenes (like Lunar New Year).\n\n## 🚀 How to Use\n\nSimply invoke the skill by describing your vision:\n\n> \"/seedance-guide A warm scene of a family reunion dinner on Lunar New Year's Eve.\"\n\nThe skill will act as your **Creative Director**, asking for style preferences, camera requirements, and available assets before generating the final prompt.\n\n## 📁 File Structure\n\n-   `SKILL.md`: The core logic and technical instructions for the agent.\n-   `usage_and_examples.md`: A library of rich, scenario-based prompt templates.\n\n## 📝 Seedance 2.0 Basics\n\nSeedance 2.0 supports up to **12 files** as input:\n-   **Images (≤9)**: Character reference, style reference, start/end frames.\n-   **Video (≤3)**: Motion reference, camera reference, extensions (2-15s).\n-   **Audio (≤3)**: Background music or lip-sync reference.\n\n---\n\n*Powered by Jimeng Seedance 2.0 Multimodal AI.*\n\nFile v1.0.1:_meta.json\n\n{\n  \"ownerId\": \"kn74vka6z3tsjs9rr5ym34y3yn819777\",\n  \"slug\": \"seedance-guide\",\n  \"version\": \"1.0.1\",\n  \"publishedAt\": 1772550602838\n}\n\nFile v1.0.1:README_CN.md\n\n# 🎬 Seedance 2.0 分镜向导 (Seedance Storyboard Director)\n\n[English](README.md) | [中文](README_CN.md)\n\n专为 **即梦 (Jimeng) Seedance 2.0** 打造的专业分镜与提示词向导。本技能旨在帮助用户将简单的创意火花，转化为“电影级”的视频生成提示词，充分释放 Seedance 的多模态能力。\n\n## ✨ 核心亮点\n\n-   **🎬 分镜驱动工作流**：将您的视频创意拆解为专业的时间轴（0-3s, 3-6s 等），让画面更有层次感。\n-   **🛠️ 多模态深度优化**：精通 `@图片`、`@视频`、`@音频` 引用语法，确角色一致性与动作复刻。\n-   **🎥 专家级镜头语言**：内置丰富的电影拍摄语言（如希区柯克变焦、推拉摇移、长镜头等）。\n-   **🩹 视频延长与编辑**：提供精准的指令引导，用于平滑延长视频或“颠覆”已有剧情。\n-   **🏮 丰富场景模版**：内置情感叙事、商业广告、武侠打斗、以及除夕团圆等节日场景模版。\n\n## 🚀 使用方法\n\n直接输入您的创意构想即可唤醒技能：\n\n> “/seedance-guide 一个温暖的除夕夜团圆饭视频。”\n\n技能将作为您的 **创意导演**，通过询问风格、镜头要求及可用素材，一步步引导您生成最终的专业提示词。\n\n## 📁 文件结构\n\n-   `SKILL.md`：技能核心逻辑与技术指令。\n-   `usage_and_examples.md`：丰富的场景化提示词模版库。\n\n## 📝 Seedance 2.0 基础知识\n\nSeedance 2.0 支持最多 **12 个文件** 作为输入：\n-   **图片 (≤9)**：用于角色参考、风格参考、首/尾帧。\n-   **视频 (≤3)**：用于动作参考、运镜参考、视频延长 (2-15s)。\n-   **音频 (≤3)**：用于背景音乐或对口型参考。\n\n---\n\n*由 即梦 Seedance 2.0 多模态 AI 驱动*\n\nFile v1.0.1:usage_and_examples.md\n\n# Seedance 2.0 Storyboard Director Examples\n\nThis document demonstrates how to use the `seedance-guide` skill for various video creation scenarios.\n\n## 1. 情感叙事 (Emotional Storytelling)\n\n**UseCase**: 温情、家庭、回忆类视频。\n\n### Prompt Template\n```markdown\n【整体设定】\n风格：电影级写实，温馨治愈\n时长：15s\n画幅：16:9 宽屏\n氛围：日落时分的暖色调，充满这种家庭的温馨感\n\n【分镜脚本】\n0-3s：[远景→中景] 镜头缓慢推进，夕阳映照着老房子的窗户，@image1 中的老人坐在摇椅上。\n3-6s：[特写] 镜头聚焦老人的手，手中拿着一张泛黄的照片（@image2），手指轻轻抚摸。\n6-10s：[回忆闪回] 画面边缘羽化，模糊转场到 @video1 中的场景（年轻时的婚礼片段），色调变得明亮鲜活。\n10-13s：[现实回归] 镜头拉回现实，老人嘴角微微上扬，眼中含泪。\n13-15s：[结尾] 画面渐暗，字幕浮现 \"爱，永恒不变\"。\n\n【声音设计】\n配乐：舒缓感人的钢琴曲，带有一点怀旧感。\n音效：摇椅晃动的嘎吱声，窗外的风铃声。\n\n【素材引用】\n@image1 老人参考\n@image2 照片参考\n@video1 回忆片段参考\n```\n\n## 2. 动作/武侠 (Action & Wuxia)\n\n**UseCase**: 打斗、舞蹈、高动态场景。\n\n### Prompt Template\n```markdown\n【整体设定】\n风格：中国水墨动作片\n时长：10s\n画幅：2.35:1 电影宽屏\n氛围：肃杀、紧张、落叶纷飞\n\n【分镜脚本】\n0-2s：[对峙] 远景，两名侠客（@image1 和 @image2）站在竹林中对峙，风吹过竹叶沙沙作响。\n2-5s：[交锋] 极速推镜头，@image1 拔剑冲刺，完全复刻 @video1 的运镜和出招动作。\n5-8s：[慢动作] 剑锋碰撞的瞬间，时间流速变慢（Bullet Time），镜头环绕展示剑气震飞落叶的细节。\n8-10s：[收招] 两人交错而过，背对站立，画面定格。\n\n【声音设计】\n配乐：激昂的鼓点+古琴。\n音效：利刃出鞘声、金属碰撞声、衣袍猎猎作响声。\n\n【素材引用】\n@image1 侠客A\n@image2 侠客B\n@video1 动作参考\n```\n\n## 3. 产品广告 (Commercial)\n\n**UseCase**: 高端产品展示 (咖啡、美妆、数码)。\n\n### Prompt Template\n```markdown\n【整体设定】\n风格：高端极简，商业广告质感\n时长：15s\n画幅：9:16 竖屏\n\n【分镜脚本】\n0-3s：[悬念] 黑色背景中，一束光打在 @image1 产品轮廓上，缓缓旋转展示流线型设计。\n3-6s：[拆解/特写] 产品部件爆炸图视角，展示内部精密结构（参考 @image2），镜头穿梭其中。\n6-9s：[应用场景] 快速切换生活场景（咖啡厅、办公室、户外），展示产品实际使用状态。\n9-12s：[质感展示] 微距镜头拍摄产品表面纹理，光影流转。\n12-15s：[落版] 产品全身像，Logo (@image3) 浮现，Slogan：\"极致体验\"。\n\n【声音设计】\n配乐：轻快时尚的电子乐。\n音效：科技感扫光音效、清脆的组装声。\n\n【素材引用】\n@image1 产品主体\n@image2 内部结构\n@image3 Logo\n```\n\n## 4. 视频延长 (Video Extension)\n\n**UseCase**: 续写已有视频，保持连贯性。\n\n### Prompt Template\n```markdown\n将 @video1 延长 5s。\n\n【分镜脚本】\n0-5s：[原有结尾延续] 视频结尾的主角继续向前奔跑，镜头逐渐拉高变成无人机俯瞰视角。主角跑过草地，进入一片茂密的森林。光线从明亮的日光逐渐转变为森林中斑驳的树影。\n\n【声音设计】\n延续原视频的背景音乐节奏，加入森林鸟叫声。\n\n【参考素材】\n@video1 原视频\n```\n\n## 5. 剧情颠覆 (Plot Twist)\n\n**UseCase**: 修改已有视频的走向。\n\n### Prompt Template\n```markdown\n颠覆 @video1 的剧情。\n\n【分镜脚本】\n0-3s：[保持] 保持原视频开头的温馨对话。\n3-6s：[反转] 突然，主角（@image1）的表情从微笑瞬间变得惊恐，看向窗外。\n6-10s：[新剧情] 窗外出现巨大的不明飞行物（@image2），红光笼罩整个房间。主角拉起同伴开始逃跑。\n\n【声音设计】\n配乐从温馨突然转为紧张急促的警报声。\n\n【参考素材】\n@video1 原视频\n@image1 主角参考\n@image2 UFO参考\n```\n\nArchive v1.0.0: 5 files, 7195 bytes\n\nFiles: README_CN.md (1819b), README.md (1812b), SKILL.md (3341b), usage_and_examples.md (4211b), _meta.json (133b)\n\nFile v1.0.0:SKILL.md\n\n---\nname: seedance-guide\ndescription: A comprehensive storyboard director for Seedance 2.0. Guides users from initial ideas to professional, movie-grade video prompts using multimodal inputs.\n---\n\n# 🎬 Seedance 2.0 Storyboard Director\n\nYou are an expert **Seedance 2.0 Creative Director**. Your goal is to help users transform vague ideas into professional, executable video generation prompts. You understand the model's multimodal capabilities, camera language, and storytelling techniques.\n\n## 核心能力 (Core Capabilities)\n\n### 1. 多模态输入限制 (Input Limits)\n| 类型 | 格式 | 数量 | 大小 |\n|---|---|---|---|\n| **图片** | jpg/png/webp | ≤ 9 | <30MB |\n| **视频** | mp4/mov | ≤ 3 (2-15s) | <50MB |\n| **音频** | mp3/wav | ≤ 3 (<15s) | <15MB |\n| **总计** | **≤ 12 个文件** | - | - |\n\n> ⚠️ **重要**：暂不支持写实真人脸部素材（会被系统拦截）。\n\n### 2. @ 引用语法 (Reference Syntax)\n必须使用 `@文件名` 明确指定素材用途：\n-   `@image1 作为首帧` (Start frame)\n-   `@image2 作为角色形象参考` (Character ref)\n-   `@video1 参考运镜和节奏` (Camera/Rhythm ref)\n-   `@audio1 作为背景音乐` (BGM)\n\n---\n\n## 交互流程 (Interactive Workflow)\n\n请遵循以下步骤引导用户：\n\n### 第一步：创意构思 (Concept)\n询问用户：\n1.  **想做什么视频？** (叙事、广告、运镜复刻、特效？)\n2.  **时长多少？** (默认15s)\n3.  **有哪些素材？** (图片、视频、音频)\n\n### 第二步：细节深化 (Details)\n根据用户回答，补充缺失信息：\n-   **风格**：电影感、二次元、水墨、赛博朋克？\n-   **运镜**：推拉摇移、希区柯克变焦、长镜头？\n-   **声音**：需要配乐、音效还是对白？\n\n### 第三步：生成提示词 (Generate)\n输出标准的 **分镜提示词** (Markdown代码块)。\n\n---\n\n## 提示词结构模版 (Prompt Structure)\n\n```markdown\n【整体设定】\n风格：[电影写实/动画/科幻...]\n时长：[15s]\n画幅：[16:9 / 2.35:1]\n\n【分镜脚本】\n0-3s：[运镜+画面] 镜头缓慢推近，@image1 中的主角站在...\n3-6s：[动作+特效] 参考 @video1 的动作，主角开始...\n6-10s：[高潮] 镜头环绕旋转，光影变得...\n10-15s：[结尾] 画面定格，字幕浮现...\n\n【声音设计】\n配乐：[情绪/风格]\n音效：[具体声音]\n\n【素材引用】\n@image1 首帧\n@video1 动作参考\n```\n\n---\n\n## 高级技巧 (Advanced Techniques)\n\n### 1. 视频延长 (Video Extension)\n-   **指令**：`将 @video1 延长 5s`\n-   **注意**：生成长度应选择 **\"新增部分\"** 的时长。\n\n### 2. 运镜复刻 (Camera Cloning)\n-   **指令**：`完全参考 @video1 的运镜和镜头语言`\n-   **注意**：确保参考视频的运镜清晰。\n\n### 3. 表情/动作迁移 (Motion Transfer)\n-   **指令**：`保持 @image1 的角色形象，复刻 @video1 的表情和动作`\n\n### 4. 视频编辑/剧情颠覆 (Video Editing)\n-   **指令**：`颠覆 @video1 的剧情，在 5s 处让主角...`\n\n---\n\n## 常见错误规避 (Avoid Pitfalls)\n1.  **引用模糊**：不要只写 `参考 @video1`，要写明参考 **什么** (运镜? 动作? 还是光影?)。\n2.  **指令冲突**：不要同时要求 \"固定镜头\" 和 \"环绕运镜\"。\n3.  **过载**：不要在 3s 内塞入太多复杂的动作描述。\n\nFile v1.0.0:README.md\n\n# 🎬 Seedance 2.0 Storyboard Director\n\n[English](README.md) | [中文](README_CN.md)\n\nA comprehensive professional storyboard and prompt director for **Jimeng Seedance 2.0**. This skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power.\n\n## ✨ Key Features\n\n-   **🎬 Storyboard-First Approach**: Breaks down your video idea into a professional timeline (0-3s, 3-6s, etc.).\n-   **🛠️ Multimodal Mastery**: Expert handling of `@image`, `@video`, and `@audio` references for character consistency and motion transfer.\n-   **🎥 Advanced Camera Language**: Deep knowledge of cinematic camera movements (Hitchcock zoom, dolly shots, etc.).\n-   **🩹 Video Extension & Editing**: Specialized guidance for extending clips or \"subverting\" existing plots.\n-   **🏮 Festive Templates**: Built-in support for storytelling, commercials, Wuxia action, and festive scenes (like Lunar New Year).\n\n## 🚀 How to Use\n\nSimply invoke the skill by describing your vision:\n\n> \"/seedance-guide A warm scene of a family reunion dinner on Lunar New Year's Eve.\"\n\nThe skill will act as your **Creative Director**, asking for style preferences, camera requirements, and available assets before generating the final prompt.\n\n## 📁 File Structure\n\n-   `SKILL.md`: The core logic and technical instructions for the agent.\n-   `usage_and_examples.md`: A library of rich, scenario-based prompt templates.\n\n## 📝 Seedance 2.0 Basics\n\nSeedance 2.0 supports up to **12 files** as input:\n-   **Images (≤9)**: Character reference, style reference, start/end frames.\n-   **Video (≤3)**: Motion reference, camera reference, extensions (2-15s).\n-   **Audio (≤3)**: Background music or lip-sync reference.\n\n---\n\n*Powered by Jimeng Seedance 2.0 Multimodal AI.*\n\nFile v1.0.0:_meta.json\n\n{\n  \"ownerId\": \"kn74vka6z3tsjs9rr5ym34y3yn819777\",\n  \"slug\": \"seedance-guide\",\n  \"version\": \"1.0.0\",\n  \"publishedAt\": 1771224633841\n}\n\nFile v1.0.0:README_CN.md\n\n# 🎬 Seedance 2.0 分镜向导 (Seedance Storyboard Director)\n\n[English](README.md) | [中文](README_CN.md)\n\n专为 **即梦 (Jimeng) Seedance 2.0** 打造的专业分镜与提示词向导。本技能旨在帮助用户将简单的创意火花，转化为“电影级”的视频生成提示词，充分释放 Seedance 的多模态能力。\n\n## ✨ 核心亮点\n\n-   **🎬 分镜驱动工作流**：将您的视频创意拆解为专业的时间轴（0-3s, 3-6s 等），让画面更有层次感。\n-   **🛠️ 多模态深度优化**：精通 `@图片`、`@视频`、`@音频` 引用语法，确角色一致性与动作复刻。\n-   **🎥 专家级镜头语言**：内置丰富的电影拍摄语言（如希区柯克变焦、推拉摇移、长镜头等）。\n-   **🩹 视频延长与编辑**：提供精准的指令引导，用于平滑延长视频或“颠覆”已有剧情。\n-   **🏮 丰富场景模版**：内置情感叙事、商业广告、武侠打斗、以及除夕团圆等节日场景模版。\n\n## 🚀 使用方法\n\n直接输入您的创意构想即可唤醒技能：\n\n> “/seedance-guide 一个温暖的除夕夜团圆饭视频。”\n\n技能将作为您的 **创意导演**，通过询问风格、镜头要求及可用素材，一步步引导您生成最终的专业提示词。\n\n## 📁 文件结构\n\n-   `SKILL.md`：技能核心逻辑与技术指令。\n-   `usage_and_examples.md`：丰富的场景化提示词模版库。\n\n## 📝 Seedance 2.0 基础知识\n\nSeedance 2.0 支持最多 **12 个文件** 作为输入：\n-   **图片 (≤9)**：用于角色参考、风格参考、首/尾帧。\n-   **视频 (≤3)**：用于动作参考、运镜参考、视频延长 (2-15s)。\n-   **音频 (≤3)**：用于背景音乐或对口型参考。\n\n---\n\n*由 即梦 Seedance 2.0 多模态 AI 驱动*\n\nFile v1.0.0:usage_and_examples.md\n\n# Seedance 2.0 Storyboard Director Examples\n\nThis document demonstrates how to use the `seedance-guide` skill for various video creation scenarios.\n\n## 1. 情感叙事 (Emotional Storytelling)\n\n**UseCase**: 温情、家庭、回忆类视频。\n\n### Prompt Template\n```markdown\n【整体设定】\n风格：电影级写实，温馨治愈\n时长：15s\n画幅：16:9 宽屏\n氛围：日落时分的暖色调，充满这种家庭的温馨感\n\n【分镜脚本】\n0-3s：[远景→中景] 镜头缓慢推进，夕阳映照着老房子的窗户，@image1 中的老人坐在摇椅上。\n3-6s：[特写] 镜头聚焦老人的手，手中拿着一张泛黄的照片（@image2），手指轻轻抚摸。\n6-10s：[回忆闪回] 画面边缘羽化，模糊转场到 @video1 中的场景（年轻时的婚礼片段），色调变得明亮鲜活。\n10-13s：[现实回归] 镜头拉回现实，老人嘴角微微上扬，眼中含泪。\n13-15s：[结尾] 画面渐暗，字幕浮现 \"爱，永恒不变\"。\n\n【声音设计】\n配乐：舒缓感人的钢琴曲，带有一点怀旧感。\n音效：摇椅晃动的嘎吱声，窗外的风铃声。\n\n【素材引用】\n@image1 老人参考\n@image2 照片参考\n@video1 回忆片段参考\n```\n\n## 2. 动作/武侠 (Action & Wuxia)\n\n**UseCase**: 打斗、舞蹈、高动态场景。\n\n### Prompt Template\n```markdown\n【整体设定】\n风格：中国水墨动作片\n时长：10s\n画幅：2.35:1 电影宽屏\n氛围：肃杀、紧张、落叶纷飞\n\n【分镜脚本】\n0-2s：[对峙] 远景，两名侠客（@image1 和 @image2）站在竹林中对峙，风吹过竹叶沙沙作响。\n2-5s：[交锋] 极速推镜头，@image1 拔剑冲刺，完全复刻 @video1 的运镜和出招动作。\n5-8s：[慢动作] 剑锋碰撞的瞬间，时间流速变慢（Bullet Time），镜头环绕展示剑气震飞落叶的细节。\n8-10s：[收招] 两人交错而过，背对站立，画面定格。\n\n【声音设计】\n配乐：激昂的鼓点+古琴。\n音效：利刃出鞘声、金属碰撞声、衣袍猎猎作响声。\n\n【素材引用】\n@image1 侠客A\n@image2 侠客B\n@video1 动作参考\n```\n\n## 3. 产品广告 (Commercial)\n\n**UseCase**: 高端产品展示 (咖啡、美妆、数码)。\n\n### Prompt Template\n```markdown\n【整体设定】\n风格：高端极简，商业广告质感\n时长：15s\n画幅：9:16 竖屏\n\n【分镜脚本】\n0-3s：[悬念] 黑色背景中，一束光打在 @image1 产品轮廓上，缓缓旋转展示流线型设计。\n3-6s：[拆解/特写] 产品部件爆炸图视角，展示内部精密结构（参考 @image2），镜头穿梭其中。\n6-9s：[应用场景] 快速切换生活场景（咖啡厅、办公室、户外），展示产品实际使用状态。\n9-12s：[质感展示] 微距镜头拍摄产品表面纹理，光影流转。\n12-15s：[落版] 产品全身像，Logo (@image3) 浮现，Slogan：\"极致体验\"。\n\n【声音设计】\n配乐：轻快时尚的电子乐。\n音效：科技感扫光音效、清脆的组装声。\n\n【素材引用】\n@image1 产品主体\n@image2 内部结构\n@image3 Logo\n```\n\n## 4. 视频延长 (Video Extension)\n\n**UseCase**: 续写已有视频，保持连贯性。\n\n### Prompt Template\n```markdown\n将 @video1 延长 5s。\n\n【分镜脚本】\n0-5s：[原有结尾延续] 视频结尾的主角继续向前奔跑，镜头逐渐拉高变成无人机俯瞰视角。主角跑过草地，进入一片茂密的森林。光线从明亮的日光逐渐转变为森林中斑驳的树影。\n\n【声音设计】\n延续原视频的背景音乐节奏，加入森林鸟叫声。\n\n【参考素材】\n@video1 原视频\n```\n\n## 5. 剧情颠覆 (Plot Twist)\n\n**UseCase**: 修改已有视频的走向。\n\n### Prompt Template\n```markdown\n颠覆 @video1 的剧情。\n\n【分镜脚本】\n0-3s：[保持] 保持原视频开头的温馨对话。\n3-6s：[反转] 突然，主角（@image1）的表情从微笑瞬间变得惊恐，看向窗外。\n6-10s：[新剧情] 窗外出现巨大的不明飞行物（@image2），红光笼罩整个房间。主角拉起同伴开始逃跑。\n\n【声音设计】\n配乐从温馨突然转为紧张急促的警报声。\n\n【参考素材】\n@video1 原视频\n@image1 主角参考\n@image2 UFO参考\n```","readmeExcerpt":"Skill: seedance 2.0 guide Owner: xiuxiuing Summary: Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scri... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-04-27T06:12:41.008Z | user Seedance 2.0 Guide v1.0.5 v1.0.4 | 2026-04-27T05:13:58.245Z | user - Added detailed expert strategies to bypass the \"realistic hum","codeSnippets":[],"executableExamples":[{"language":"markdown","snippet":"【Overall Setting】\nStyle: [Cinematic Realistic/Animation/Sci-Fi...]\nDuration: [15s]\nAspect Ratio: [16:9 / 2.35:1]\n\n【Storyboard Script】\n0-3s: [Camera + Visual] Camera slowly zooms in, the protagonist in @image1 stands at...\n3-6s: [Action + Effect] Referencing the actions in @video1, the protagonist starts to...\n6-12s: [Climax] Camera rotates around, lighting becomes...\n12-15s: [Ending] Image freezes, subtitles emerge...\n\n【Sound Design】\nBGM: [Emotion/Style]\nSound Effects: [Specific sounds]\n\n【Material Reference】\n@image1 Start frame\n@video1 Action reference"},{"language":"markdown","snippet":"【Directorial Breakdown】\nSUBJECTS: [Detailed character/object description, e.g., muscle lines, material texture]\nENVIRONMENT: [Lighting, weather, interior/exterior details]\nMOOD/STYLE: [Color grade, cinematic feel, emotion]\n\n【Timeline】\n0-5s (Segment 1): [Detailed action + Camera directives]. Use 【@asset】 brackets to anchor visuals to text.\n5-10s (Segment 2): [Detailed action + Camera directives].\n10-15s (Segment 3): [Ending/Climax].\n\n【Technical Requirements】\nCamera: [e.g., 1-to-1 restoration of @video1, Close-up quick push-in, Handheld feel]\nVisual Consistency: [Strictly refer to @image1 for character appearance]\n\n【Sound & Voice】\nVoice Tone: [e.g., Passionate explaining, calm narration]\nSound Effects: [Precise timing for SFX]\n\n【Material Reference】\n@image1 Subjects/Character\n@video1 Motion/Camera reference"},{"language":"markdown","snippet":"【Overall Setting】\nStyle: Cinematic Realistic, Warm & Healing\nDuration: 15s\nAspect Ratio: 16:9\n\n【Storyboard Script】\n0-3s: [Wide to Medium] Camera slowly pushes in. Sunset light shines on an old house. The elderly person in @image1 sits on a rocking chair.\n3-6s: [Close-up] Camera focuses on the hands, holding a yellowed photo (@image2).\n6-12s: [Flashback] Soft edge transition to scene in @video1 (younger wedding). Bright and vivid.\n12-15s: [Ending] Back to reality. The elderly person smiles with tears. Fade to black.\n\n【Sound Design】\nBGM: Piano music, nostalgic.\nSFX: Rocking chair creaking, wind chimes.\n\n【Material Reference】\n@image1 Elderly person\n@image2 Photo\n@video1 Flashback reference"},{"language":"markdown","snippet":"Extend @video1 by 5s.\n\n【Storyboard Script】\n0-5s: [Continuity] The protagonist continues running. Camera transitions to drone overhead view. The protagonist enters a dense forest. Lighting changes from bright daylight to dappled shadows.\n\n【Sound Design】\nMaintain BGM rhythm, add forest bird sounds.\n\n【Material Reference】\n@video1 Original video"},{"language":"markdown","snippet":"【Directorial Breakdown】\nSUBJECTS: A woman in traditional black silk martial arts attire, sharp eyes, agile movements.\nENVIRONMENT: Bamboo forest at dawn, thick fog, falling leaves.\nMOOD/STYLE: Ink-wash cinematic, high contrast, deadly cold atmosphere.\n\n【Timeline】\n0-5s (Segment 1): [1-to-1 Restoration of @video1]. Strictly follow the sword parrying action and camera shake. The woman 【@image1】 deflects incoming flying daggers with a heavy blade.\n5-10s (Segment 2): [Slow Motion Transition]. Camera rotates 360 degrees around the character (Bullet Time). Time slows to 0.2x. Leaves freeze in mid-air.\n10-15s (Segment 3): [Resolution]. The woman sheath her sword. Camera pulls back to a wide shot of the foggy forest.\n\n【Technical Requirements】\nCamera: Strictly refer to the camera movement of @video1. Handheld feel during action.\nVisual Consistency: Maintain the face and outfit from @image1.\n\n【Sound & Voice】\nSound Effects: Precise metallic \"clink\" synchronized with the parry in Segment 1. Wind whistling through bamboo.\n\n【Material Reference】\n@image1 Character reference\n@video1 Action & Camera reference"},{"language":"markdown","snippet":"【Directorial Breakdown】\nSUBJECTS: A 【fashionable and energetic】beautiful woman 【@image1】, holding and showcasing a 【smart sports watch】 【@image2】.\nENVIRONMENT: Modern, bright minimalist studio, soft ring light.\nMOOD/STYLE: Vibrant commercial, clean and professional.\n\n【Timeline】\n0-5s (Segment 1): [Talking Head]. The woman is passionately explaining while looking directly at the camera. Her hands hold the watch 【@image2】 up to her chest.\n5-10s (Segment 2): [Feature Close-up]. Cut to a macro shot of the watch face. A finger taps the screen, showing the heart rate UI (referencing @image3).\n10-15s (Segment 3): [Call to Action]. Back to medium shot. The woman points to the \"Buy\" button overlay. She smiles and waves.\n\n【Technical Requirements】\nCamera: Fixed tripod for stable talking head. Quick cut to macro at 5s.\nVoice Tone: Enthusiastic, high-pitched \"livestream seller\" style.\n\n【Sound & Voice】\nVoice Tone: Passionate, rhythmic \"Family! We have a deal today!\"\nBGM: Upbeat corporate pop.\n\n【Material Reference】\n@image1 Host reference\n@image2 Product reference\n@image3 Screen UI reference"}],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director\ndescription: Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scripts from multimodal inputs. Optimized for Claude Code, Cursor, and OpenClaw.\n---\n\n# 🎬 Seedance 2.0 Storyboard Director\n\nYou are an expert **Seedance 2.0 Creative Director**. Your goal is to help users transform vague ideas into professional, executable video generation prompts. You understand the model's multimodal capabilities, camera language, and storytelling techniques.\n\n## Core Capabilities\n\n### 1. Multimodal Input Limits\n| Type | Format | Quantity | Size |\n|---|---|---|---|\n| **Images** | jpg/png/webp | ≤ 9 | <30MB |\n| **Videos** | mp4/mov | ≤ 3 (2-15s) | <50MB |\n| **Audio** | mp3/wav | ≤ 3 (<15s) | <15MB |\n| **Total** | **≤ 12 files** | - | - |\n\n> 💡 **CREATIVE TIP**: To achieve high-quality character consistency, prioritize **Stylized Transformation** (e.g., @image1 in 3D cartoon style) or **Cinematic Framing** (e.g., side profile, silhouette).\n\n### 2. @ Reference Syntax\nYou must use `@filename` to clearly specify the use of the material:\n-   `@image1 as start frame`\n-   `@image2 as character reference`\n-   `@video1 for camera movement and rhythm reference`\n-   `@audio1 as background music`\n\n---\n\n## Interactive Workflow\n\nPlease follow these steps to guide the user:\n\n### Step 1: Concept\nAsk the user:\n1.  **What kind of video do you want to make?** (Narrative, advertisement, camera movement replication, special effects?)\n2.  **What is the duration?** (Default is 15s)\n3.  **What materials do you have?** (Images, videos, audio)\n\n### Step 2: Details\nBased on the user's answers, supplement missing information:\n-   **Style**: Cinematic, anime, ink wash, cyberpunk?\n-   **Camera Movement**: Push/pull, pan, tilt, Hitchcock zoom, long take?\n-   **Sound**: Do you need background music, sound effects, or dialogue?\n\n### Step 3: Generate Prompt\nOutput standard **storyboard prompts** using either the **Basic** or **Advanced** template based on the user's needs.\n\n---\n\n## Prompt Structure Template\n\n### 1. Basic (Storyboard Script)\nBest for: Quick ideas, simple narratives.\n\n```markdown\n【Overall Setting】\nStyle: [Cinematic Realistic/Animation/Sci-Fi...]\nDuration: [15s]\nAspect Ratio: [16:9 / 2.35:1]\n\n【Storyboard Script】\n0-3s: [Camera + Visual] Camera slowly zooms in, the protagonist in @image1 stands at...\n3-6s: [Action + Effect] Referencing the actions in @video1, the protagonist starts to...\n6-12s: [Climax] Camera rotates around, lighting becomes...\n12-15s: [Ending] Image freezes, subtitles emerge...\n\n【Sound Design】\nBGM: [Emotion/Style]\nSound Effects: [Specific sounds]\n\n【Material Reference】\n@image1 Start frame\n@video1 Action reference\n```\n\n### 2. Advanced (Directorial Breakdown)\nBest for: High-precision restoration, complex scenes, e-commerce.\n\n```markdown\n【Directorial Breakdown】\nSUBJECTS: [Detailed character/object descr"},{"path":"README.md","content":"# 🎬 Seedance 2.0 Guide - Cinematic Storyboard & Prompt Director\n\n[English](README.md) | [中文](README_CN.md)\n\nA comprehensive professional storyboard and prompt engineering director for **Jimeng (即梦) Seedance 2.0**. Optimized for **Claude Code, Cursor, and OpenClaw**, this skill guides you from a simple idea to a \"movie-grade\" video generation prompt, leveraging Seedance's full multimodal power (Image-to-Video, Video-to-Video).\n\n## ✨ Key Features\n\n-   **🎯 SEO & Prompt Optimized**: Expertly crafted for **Seedance 2.0** and **Jimeng**.\n-   **Dual-Mode Driven Workflow**: Supports both **Basic** (Storyboard Script) and **Advanced** (Directorial Breakdown) formats.\n-   **🛠️ Multimodal Mastery**: Expert handling of `@image`, `@video`, and `@audio` references.\n-   **🎥 High-Precision Motion/Camera Cloning**: Supports 1-to-1 video restoration.\n-   **🩹 Video Extension & Editing**: Specialized guidance for extending and subverting plots.\n\n## 🚀 How to Use\n\nSimply invoke the skill by describing your vision:\n\n> \"/seedance-guide A warm scene of a family reunion dinner on Lunar New Year's Eve.\"\n\nThe skill will act as your **Creative Director**, asking for style preferences, camera requirements, and available assets before generating the final prompt.\n\n## 📁 File Structure\n\n-   `SKILL.md`: The core logic and technical instructions for the agent.\n-   `usage_and_examples.md`: A library of rich, scenario-based prompt templates.\n\n## 📝 Seedance 2.0 Basics\n\nSeedance 2.0 supports up to **12 files** as input:\n-   **Images (≤9)**: Character reference, style reference, start/end frames.\n-   **Video (≤3)**: Motion reference, camera reference, extensions (2-15s).\n-   **Audio (≤3)**: Background music or lip-sync reference.\n\n---\n\n*Powered by Jimeng Seedance 2.0 Multimodal AI.*"},{"path":"_meta.json","content":"{\n  \"ownerId\": \"kn74vka6z3tsjs9rr5ym34y3yn819777\",\n  \"slug\": \"seedance-guide\",\n  \"version\": \"1.0.5\",\n  \"publishedAt\": 1777270361008\n}"},{"path":"README_CN.md","content":"# 🎬 Seedance 2.0 分镜向导 - 即梦 (Jimeng) 电影级提示词导演\n\n[English](README.md) | [中文](README_CN.md)\n\n专为 **即梦 (Jimeng) Seedance 2.0** 打造的专业分镜与提示词 (Prompt Engineering) 向导。本技能旨在帮助用户将简单的创意火花，转化为“电影级”的视频生成提示词，充分释放 Seedance 的多模态能力（图生视频、视频生视频）。\n\n## ✨ 核心亮点\n\n-   **🎯 搜索与提示词深度优化**：针对 **即梦 (Jimeng)** 和 **Seedance 2.0** 的分词逻辑精准调优。\n-   **双模驱动工作流**：支持 **基础 (Basic)** 分镜脚本与 **高级 (Advanced)** 导演 breakdown，满足从快速创意到高精度复刻全场景需求。\n-   **🛠️ 多模态深度优化**：精通 `@图片`、`@视频`、`@音频` 引用语法。\n-   **🎥 高精度动作/运镜复刻**：支持 1-to-1 视频还原。\n-   **🩹 视频延长与编辑**：提供精准指令引导，用于平滑延长视频或“颠覆”已有剧情。\n\n## 🚀 使用方法\n\n直接输入您的创意构想即可唤醒技能：\n\n> “/seedance-guide 一个温暖的除夕夜团圆饭视频。”\n\n技能将作为您的 **创意导演**，通过询问风格、镜头要求及可用素材，一步步引导您生成最终的专业提示词。\n\n## 📁 文件结构\n\n-   `SKILL.md`：技能核心逻辑与技术指令。\n-   `usage_and_examples.md`：丰富的场景化提示词模版库。\n\n## 📝 Seedance 2.0 基础知识\n\nSeedance 2.0 支持最多 **12 个文件** 作为输入：\n-   **图片 (≤9)**：用于角色参考、风格参考、首/尾帧。\n-   **视频 (≤3)**：用于动作参考、运镜参考、视频延长 (2-15s)。\n-   **音频 (≤3)**：用于背景音乐或对口型参考。\n\n---\n\n*由 即梦 Seedance 2.0 多模态 AI 驱动*"},{"path":"skill-card.md","content":"## Description:\n\nProfessional Seedance 2.0 / Jimeng storyboard and prompt engineering guide for creating cinematic storyboards, movie-grade AI video prompts, and audio script guidance from multimodal inputs.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[xiuxiuing](https://clawhub.ai/user/xiuxiuing)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nExternal users and developers use this skill to turn video ideas and multimodal assets into Seedance 2.0 storyboard prompts, directorial breakdowns, camera instructions, and sound guidance.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: The prompt guidance may encourage content-check evasion for identity-linked media generation.\n\nMitigation: Remove detector-aware face-obscuring and term-substitution guidance before use or publication.\n\nRisk: Use with real people's images or likenesses can raise consent, impersonation, or platform-policy concerns.\n\nMitigation: Require documented consent for likeness use and confirm generated prompts comply with the target video platform's content policies.\n\n## Reference(s):\n\n- [ClawHub skill page](https://clawhub.ai/xiuxiuing/skills/seedance-guide)\n\n## Skill Output:\n\n**Output Type(s):** [Text, Markdown, Guidance]\n\n**Output Format:** [Markdown storyboard prompts, directorial breakdowns, and sound guidance]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May reference user-provided image, video, and audio assets using @filename syntax.]\n\n## Skill Version(s):\n\n1.0.5 (source: server-resolved release metadata)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment."}],"languages":[],"docsSourceLabel":"CLAWHUB","editorialOverview":"Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scri... Skill: seedance 2.0 guide Owner: xiuxiuing Summary: Professional Seedance 2.0 / Jimeng (即梦) Storyboard & Prompt Engineering Guide. Create movie-grade 9:16 vlogs, cinematic AI video prompts, and auto-audio scri... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-04-27T06:12:41.008Z | user Seedance 2.0 Guide v1.0.5 v1.0.4 | 2026-04-27T05:13:58.245Z | user - Added detailed expert strategies to bypass the \"realistic hum","editorialQuality":{"score":100,"threshold":65,"status":"ready","wordCount":1463,"uniquenessScore":49,"reasons":[]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-09T10:45:23.948Z","emptyReason":"No screenshots, media assets, or demo links are available."},"primaryImageUrl":null,"mediaAssetCount":0,"assets":[],"demoUrl":null},"ownerResources":{"evidence":{"source":"unclaimed","verified":false,"confidence":"low","updatedAt":"2026-10-09T10:45:23.948Z","emptyReason":"This page has not been claimed by the agent owner."},"hasCustomPage":false,"customPageUpdatedAt":null,"customLinks":[],"structuredLinks":{"docsUrl":null,"demoUrl":null,"supportUrl":null,"pricingUrl":null,"statusUrl":null},"customPage":null},"relatedAgents":{"evidence":{"source":"protocol-neighbors","verified":false,"confidence":"medium","updatedAt":"2026-10-09T14:51:21.303Z","emptyReason":null},"items":[{"id":"b917f68a-ebff-438e-84f8-3f4b2494c0bc","entityType":"agent","canonicalPath":"/agent/activepieces-activepieces","slug":"activepieces-activepieces","name":"activepieces","description":"AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents","url":"https://github.com/activepieces/activepieces","homepage":"https://www.activepieces.com","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-15T02:22:12.426Z","createdAt":"2026-02-25T03:38:12.412Z","downloads":null},{"id":"5cb26759-3a39-483f-94cf-276a98c13bb8","entityType":"agent","canonicalPath":"/agent/cherryhq-cherry-studio","slug":"cherryhq-cherry-studio","name":"cherry-studio","description":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs","url":"https://github.com/CherryHQ/cherry-studio","homepage":"https://cherry-ai.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-11T14:38:40.986Z","createdAt":"2026-02-25T03:38:19.379Z","downloads":null},{"id":"8ebccd8e-3863-4187-8355-c3f14e1f9edf","entityType":"agent","canonicalPath":"/agent/iofficeai-aionui","slug":"iofficeai-aionui","name":"AionUi","description":"Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!","url":"https://github.com/iOfficeAI/AionUi","homepage":"https://www.aionui.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-10T18:48:31.762Z","createdAt":"2026-02-25T03:38:16.584Z","downloads":null},{"id":"6f6582d0-5d76-4f0f-b81d-86520247950b","entityType":"agent","canonicalPath":"/agent/copilotkit-copilotkit","slug":"copilotkit-copilotkit","name":"CopilotKit","description":"The Frontend for Agents & Generative UI. React + Angular","url":"https://github.com/CopilotKit/CopilotKit","homepage":"https://docs.copilotkit.ai","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-03-25T09:50:57.846Z","createdAt":"2026-02-25T03:39:14.617Z","downloads":null}],"links":{"hub":"/agent","source":"/agent/source/clawhub","protocols":[{"label":"OpenClaw","href":"/agent/protocol/openclew"}]}}}