# generative-media-skills - Doramagic AI Context Pack

> 定位：安装前体验与判断资产。它帮助宿主 AI 有一个好的开始，但不代表已经安装、执行或验证目标项目。

## 充分原则

- **充分原则，不是压缩原则**：AI Context Pack 应该充分到让宿主 AI 在开工前理解项目价值、能力边界、使用入口、风险和证据来源；它可以分层组织，但不以最短摘要为目标。
- **压缩策略**：只压缩噪声和重复内容，不压缩会影响判断和开工质量的上下文。

## 给宿主 AI 的使用方式

你正在读取 Doramagic 为 generative-media-skills 编译的 AI Context Pack。请把它当作开工前上下文：帮助用户理解适合谁、能做什么、如何开始、哪些必须安装后验证、风险在哪里。不要声称你已经安装、运行或执行了目标项目。

## Claim 消费规则

- **事实来源**：Repo Evidence + Claim/Evidence Graph；Human Wiki 只提供显著性、术语和叙事结构。
- **事实最低状态**：`supported`
- `supported`：可以作为项目事实使用，但回答中必须引用 claim_id 和证据路径。
- `weak`：只能作为低置信度线索，必须要求用户继续核实。
- `inferred`：只能用于风险提示或待确认问题，不能包装成项目事实。
- `unverified`：不得作为事实使用，应明确说证据不足。
- `contradicted`：必须展示冲突来源，不得替用户强行选择一个版本。

## 它最适合谁

- **正在使用 Claude/Codex/Cursor/Gemini 等宿主 AI 的开发者**：README 或插件配置提到多个宿主 AI。 证据：`README.md` Claim：`clm_0003` supported 0.86
- **希望把专业流程带进宿主 AI 的用户**：仓库包含 Skill 文档。 证据：`core/edit/SKILL.md`, `core/media/SKILL.md`, `core/platform/SKILL.md`, `library/edit/ai-clipping/SKILL.md` 等 Claim：`clm_0004` supported 0.86

## 它能做什么

- **AI Skill / Agent 指令资产库**（可做安装前预览）：项目包含可被宿主 AI 读取的 Skill 或 Agent 指令文件，可用于把专业流程带入 Claude、Codex、Cursor 等宿主。 证据：`core/edit/SKILL.md`, `core/media/SKILL.md`, `core/platform/SKILL.md`, `library/edit/ai-clipping/SKILL.md` 等 Claim：`clm_0001` supported 0.86
- **命令行启动或安装流程**（需要安装后验证）：项目文档中存在可执行命令，真实使用需要在本地或宿主环境中运行这些命令。 证据：`README.md` Claim：`clm_0002` supported 0.86

## 怎么开始

- `npm install -g muapi-cli` 证据：`README.md` Claim：`clm_0005` supported 0.86
- `pip install muapi-cli` 证据：`README.md` Claim：`clm_0006` supported 0.86
- `npx muapi-cli --help` 证据：`README.md` Claim：`clm_0007` supported 0.86
- `npx skills add SamurAIGPT/Generative-Media-Skills --all` 证据：`README.md` Claim：`clm_0008` supported 0.86, `clm_0010` supported 0.86
- `npx skills add SamurAIGPT/Generative-Media-Skills --skill muapi-media-generation` 证据：`README.md` Claim：`clm_0009` supported 0.86
- `npx skills add SamurAIGPT/Generative-Media-Skills --all -a claude-code -a cursor` 证据：`README.md` Claim：`clm_0010` supported 0.86

## 继续前判断卡

- **当前建议**：需要管理员/安全审批
- **为什么**：继续前可能涉及密钥、账号、外部服务或敏感上下文，建议先经过管理员或安全审批。

### 30 秒判断

- **现在怎么做**：需要管理员/安全审批
- **最小安全下一步**：先跑 Prompt Preview；若涉及凭证或企业环境，先审批再试装
- **先别相信**：工具权限边界不能在安装前相信。
- **继续会触碰**：命令执行、宿主 AI 配置、本地环境或项目文件

### 现在可以相信

- **适合人群线索：正在使用 Claude/Codex/Cursor/Gemini 等宿主 AI 的开发者**（supported）：有 supported claim 或项目证据支撑，但仍不等于真实安装效果。 证据：`README.md` Claim：`clm_0003` supported 0.86
- **适合人群线索：希望把专业流程带进宿主 AI 的用户**（supported）：有 supported claim 或项目证据支撑，但仍不等于真实安装效果。 证据：`core/edit/SKILL.md`, `core/media/SKILL.md`, `core/platform/SKILL.md`, `library/edit/ai-clipping/SKILL.md` 等 Claim：`clm_0004` supported 0.86
- **能力存在：AI Skill / Agent 指令资产库**（supported）：可以相信项目包含这类能力线索；是否适合你的具体任务仍要试用或安装后验证。 证据：`core/edit/SKILL.md`, `core/media/SKILL.md`, `core/platform/SKILL.md`, `library/edit/ai-clipping/SKILL.md` 等 Claim：`clm_0001` supported 0.86
- **能力存在：命令行启动或安装流程**（supported）：可以相信项目包含这类能力线索；是否适合你的具体任务仍要试用或安装后验证。 证据：`README.md` Claim：`clm_0002` supported 0.86
- **存在 Quick Start / 安装命令线索**（supported）：可以相信项目文档出现过启动或安装入口；不要因此直接在主力环境运行。 证据：`README.md` Claim：`clm_0005` supported 0.86

### 现在还不能相信

- **工具权限边界不能在安装前相信。**（unverified）：MCP/tool 类项目通常会触碰文件、网络、浏览器或外部 API，必须真实检查权限和日志。
- **真实输出质量不能在安装前相信。**（unverified）：Prompt Preview 只能展示引导方式，不能证明真实项目中的结果质量。
- **宿主 AI 版本兼容性不能在安装前相信。**（unverified）：Claude、Cursor、Codex、Gemini 等宿主加载规则和版本差异必须在真实环境验证。
- **不会污染现有宿主 AI 行为，不能直接相信。**（inferred）：Skill、plugin、AGENTS/CLAUDE/GEMINI 指令可能改变宿主 AI 的默认行为。 证据：`core/edit/SKILL.md`, `core/media/SKILL.md`, `core/platform/SKILL.md`, `library/edit/ai-clipping/SKILL.md` 等
- **可安全回滚不能默认相信。**（unverified）：除非项目明确提供卸载和恢复说明，否则必须先在隔离环境验证。
- **真实安装后是否与用户当前宿主 AI 版本兼容？**（unverified）：兼容性只能通过实际宿主环境验证。
- **项目输出质量是否满足用户具体任务？**（unverified）：安装前预览只能展示流程和边界，不能替代真实评测。
- **安装命令是否需要网络、权限或全局写入？**（unverified）：这影响企业环境和个人环境的安装风险。 证据：`README.md`

### 继续会触碰什么

- **命令执行**：包管理器、网络下载、本地插件目录、项目配置或用户主目录。 原因：运行第一条命令就可能产生环境改动；必须先判断是否值得跑。 证据：`README.md`
- **宿主 AI 配置**：Claude/Codex/Cursor/Gemini/OpenCode 等宿主的 plugin、Skill 或规则加载配置。 原因：宿主配置会改变 AI 后续工作方式，可能和用户已有规则冲突。 证据：`core/edit/SKILL.md`, `core/media/SKILL.md`, `core/platform/SKILL.md`, `library/edit/ai-clipping/SKILL.md` 等
- **本地环境或项目文件**：安装结果、插件缓存、项目配置或本地依赖目录。 原因：安装前无法证明写入范围和回滚方式，需要隔离验证。 证据：`README.md`
- **环境变量 / API Key**：项目入口文档明确出现 API key、token、secret 或账号凭证配置。 原因：如果真实安装需要凭证，应先使用测试凭证并经过权限/合规判断。 证据：`README.md`, `core/platform/SKILL.md`, `core/platform/setup.sh`, `library/edit/ai-clipping/SKILL.md` 等
- **宿主 AI 上下文**：AI Context Pack、Prompt Preview、Skill 路由、风险规则和项目事实。 原因：导入上下文会影响宿主 AI 后续判断，必须避免把未验证项包装成事实。

### 最小安全下一步

- **先跑 Prompt Preview**：用安装前交互式试用判断工作方式是否匹配，不需要授权或改环境。（适用：任何项目都适用，尤其是输出质量未知时。）
- **只在隔离目录或测试账号试装**：避免安装命令污染主力宿主 AI、真实项目或用户主目录。（适用：存在命令执行、插件配置或本地写入线索时。）
- **先备份宿主 AI 配置**：Skill、plugin、规则文件可能改变 Claude/Cursor/Codex 的默认行为。（适用：存在插件 manifest、Skill 或宿主规则入口时。）
- **不要使用真实生产凭证**：环境变量/API key 一旦进入宿主或工具链，可能产生账号和合规风险。（适用：出现 API、TOKEN、KEY、SECRET 等环境线索时。）
- **安装后只验证一个最小任务**：先验证加载、兼容、输出质量和回滚，再决定是否深用。（适用：准备从试用进入真实工作流时。）

### 退出方式

- **保留安装前状态**：记录原始宿主配置和项目状态，后续才能判断是否可恢复。
- **准备移除宿主 plugin / Skill / 规则入口**：如果试装后行为异常，可以把宿主 AI 恢复到试装前状态。
- **记录安装命令和写入路径**：没有明确卸载说明时，至少要知道哪些目录或配置需要手动清理。
- **准备撤销测试 API key 或 token**：测试凭证泄露或误用时，可以快速止损。
- **如果没有回滚路径，不进入主力环境**：不可回滚是继续前阻断项，不应靠信任或运气继续。

## 哪些只能预览

- 解释项目适合谁和能做什么
- 基于项目文档演示典型对话流程
- 帮助用户判断是否值得安装或继续研究

## 哪些必须安装后验证

- 真实安装 Skill、插件或 CLI
- 执行脚本、修改本地文件或访问外部服务
- 验证真实输出质量、性能和兼容性

## 边界与风险判断卡

- **把安装前预览误认为真实运行**：用户可能高估项目已经完成的配置、权限和兼容性验证。 处理方式：明确区分 prompt_preview_can_do 与 runtime_required。 Claim：`clm_0011` inferred 0.45
- **命令执行会修改本地环境**：安装命令可能写入用户主目录、宿主插件目录或项目配置。 处理方式：先在隔离环境或测试账号中运行。 证据：`README.md` Claim：`clm_0012` supported 0.86
- **待确认**：真实安装后是否与用户当前宿主 AI 版本兼容？。原因：兼容性只能通过实际宿主环境验证。
- **待确认**：项目输出质量是否满足用户具体任务？。原因：安装前预览只能展示流程和边界，不能替代真实评测。
- **待确认**：安装命令是否需要网络、权限或全局写入？。原因：这影响企业环境和个人环境的安装风险。

## 开工前工作上下文

### 加载顺序

- 先读取 how_to_use.host_ai_instruction，建立安装前判断资产的边界。
- 读取 claim_graph_summary，确认事实来自 Claim/Evidence Graph，而不是 Human Wiki 叙事。
- 再读取 intended_users、capabilities 和 quick_start_candidates，判断用户是否匹配。
- 需要执行具体任务时，优先查 role_skill_index，再查 evidence_index。
- 遇到真实安装、文件修改、网络访问、性能或兼容性问题时，转入 risk_card 和 boundaries.runtime_required。

### 任务路由

- **AI Skill / Agent 指令资产库**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`core/edit/SKILL.md`, `core/media/SKILL.md`, `core/platform/SKILL.md`, `library/edit/ai-clipping/SKILL.md` 等 Claim：`clm_0001` supported 0.86
- **命令行启动或安装流程**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`README.md` Claim：`clm_0002` supported 0.86

### 上下文规模

- 文件总数：89
- 重要文件覆盖：40/89
- 证据索引条目：73
- 角色 / Skill 条目：60

### 证据不足时的处理

- **missing_evidence**：说明证据不足，要求用户提供目标文件、README 段落或安装后验证记录；不要补全事实。
- **out_of_scope_request**：说明该任务超出当前 AI Context Pack 证据范围，并建议用户先查看 Human Manual 或真实安装后验证。
- **runtime_request**：给出安装前检查清单和命令来源，但不要替用户执行命令或声称已执行。
- **source_conflict**：同时展示冲突来源，标记为待核实，不要强行选择一个版本。

## Prompt Recipes

### 适配判断

- 目标：判断这个项目是否适合用户当前任务。
- 预期输出：适配结论、关键理由、证据引用、安装前可预览内容、必须安装后验证内容、下一步建议。

```text
请基于 generative-media-skills 的 AI Context Pack，先问我 3 个必要问题，然后判断它是否适合我的任务。回答必须包含：适合谁、能做什么、不能做什么、是否值得安装、证据来自哪里。所有项目事实必须引用 evidence_refs、source_paths 或 claim_id。
```

### 安装前体验

- 目标：让用户在安装前感受核心工作流，同时避免把预览包装成真实能力或营销承诺。
- 预期输出：一段带边界标签的体验剧本、安装后验证清单和谨慎建议；不含真实运行承诺或强营销表述。

```text
请把 generative-media-skills 当作安装前体验资产，而不是已安装工具或真实运行环境。

请严格输出四段：
1. 先问我 3 个必要问题。
2. 给出一段“体验剧本”：用 [安装前可预览]、[必须安装后验证]、[证据不足] 三种标签展示它可能如何引导工作流。
3. 给出安装后验证清单：列出哪些能力只有真实安装、真实宿主加载、真实项目运行后才能确认。
4. 给出谨慎建议：只能说“值得继续研究/试装”“先补充信息后再判断”或“不建议继续”，不得替项目背书。

硬性边界：
- 不要声称已经安装、运行、执行测试、修改文件或产生真实结果。
- 不要写“自动适配”“确保通过”“完美适配”“强烈建议安装”等承诺性表达。
- 如果描述安装后的工作方式，必须使用“如果安装成功且宿主正确加载 Skill，它可能会……”这种条件句。
- 体验剧本只能写成“示例台词/假设流程”：使用“可能会询问/可能会建议/可能会展示”，不要写“已写入、已生成、已通过、正在运行、正在生成”。
- Prompt Preview 不负责给安装命令；如用户准备试装，只能提示先阅读 Quick Start 和 Risk Card，并在隔离环境验证。
- 所有项目事实必须来自 supported claim、evidence_refs 或 source_paths；inferred/unverified 只能作风险或待确认项。

```

### 角色 / Skill 选择

- 目标：从项目里的角色或 Skill 中挑选最匹配的资产。
- 预期输出：候选角色或 Skill 列表，每项包含适用场景、证据路径、风险边界和是否需要安装后验证。

```text
请读取 role_skill_index，根据我的目标任务推荐 3-5 个最相关的角色或 Skill。每个推荐都要说明适用场景、可能输出、风险边界和 evidence_refs。
```

### 风险预检

- 目标：安装或引入前识别环境、权限、规则冲突和质量风险。
- 预期输出：环境、权限、依赖、许可、宿主冲突、质量风险和未知项的检查清单。

```text
请基于 risk_card、boundaries 和 quick_start_candidates，给我一份安装前风险预检清单。不要替我执行命令，只说明我应该检查什么、为什么检查、失败会有什么影响。
```

### 宿主 AI 开工指令

- 目标：把项目上下文转成一次对话开始前的宿主 AI 指令。
- 预期输出：一段边界明确、证据引用明确、适合复制给宿主 AI 的开工前指令。

```text
请基于 generative-media-skills 的 AI Context Pack，生成一段我可以粘贴给宿主 AI 的开工前指令。这段指令必须遵守 not_runtime=true，不能声称项目已经安装、运行或产生真实结果。
```


## 角色 / Skill 索引

- 共索引 60 个角色 / Skill / 项目文档条目。

- **muapi-media-editing**（skill）：Edit and enhance images and videos with AI via muapi.ai — prompt-based editing, upscaling, background removal, face swap, lipsync, video effects, and more 激活提示：当用户任务与“muapi-media-editing”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`core/edit/SKILL.md`
- **muapi-media-generation**（skill）：Generate AI images, videos, music, and audio from the terminal via muapi.ai — supports 100+ models including Flux, Midjourney v7, Kling 3.0, Veo3, and Suno V5 激活提示：当用户任务与“muapi-media-generation”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`core/media/SKILL.md`
- **muapi-platform**（skill）：Setup and utility scripts for muapi.ai — configure API keys, test connectivity, and poll for async generation results 激活提示：当用户任务与“muapi-platform”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`core/platform/SKILL.md`
- **muapi-ai-clipping**（skill）：Turn a long video into N viral-ready short clips with a single managed API call. Wraps muapi.ai's /ai-clipping endpoint, which handles transcription, highlight ranking through a virality framework hook / emotional peak / opinion bomb / revelation / conflict / quotable / story peak / practical value , overlap dedupe, and vertical face-tracking auto-crop server-side. No local Whisper, no local LLM, no GPU. 激活提示：当用户任务与“muapi-ai-clipping”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/edit/ai-clipping/SKILL.md`
- **muapi-3d-logo-animation**（skill）：Transform a 2D logo into a premium 3D version and animate it with professional cinematic effects. 激活提示：当用户任务与“muapi-3d-logo-animation”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/3d-logo-animation/SKILL.md`
- **muapi-ai-fight-scene**（skill）：Generate a high-cut-density action / fight scene by first composing a 16-cell storyboard image, then driving Seedance 2.0 image-to-video off that storyboard. Stacks GPT-Image-2 character sheet + storyboard , Nano-Banana-2 environment concept , and Seedance 2.0 i2v. 激活提示：当用户任务与“muapi-ai-fight-scene”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/ai-fight-scene/SKILL.md`
- **muapi-animal-video-generator**（skill）：Create a hilarious and ultra-realistic video of an anthropomorphic animal acting like a human vlogger in a real-world setting. 激活提示：当用户任务与“muapi-animal-video-generator”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/animal-video-generator/SKILL.md`
- **muapi-award-ceremony-video**（skill）：Generate a 15-second cinematic awards-ceremony video — a host announces a winner from the stage, a spotlight finds them in the crowd, they walk up to the podium, receive the award, and the LED display reveals their name and "THE BEST ACTOR". 激活提示：当用户任务与“muapi-award-ceremony-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/award-ceremony-video/SKILL.md`
- **muapi-cartoon-dance-animation**（skill）：Convert a photo of a person into a Pixar-style 3D cartoon character, then animate it using a reference dance or motion video. 激活提示：当用户任务与“muapi-cartoon-dance-animation”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/cartoon-dance-animation/SKILL.md`
- **muapi-character-story-video**（skill）：Create a multi-part animated story video by first establishing a consistent character and then generating sequential scenes and animating them. 激活提示：当用户任务与“muapi-character-story-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/character-story-video/SKILL.md`
- **muapi-cinema-director**（skill）：Direct high-fidelity cinematic video with AI — translates creative intent into technical cinematographic directives for Veo3, Kling, and Luma video models via muapi.ai 激活提示：当用户任务与“muapi-cinema-director”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/cinema-director/SKILL.md`
- **muapi-drone-style-video**（skill）：Generate aerial drone-perspective footage — sweeping bird's-eye views, orbit shots, and flyover sequences for landscapes, architecture, and events. 激活提示：当用户任务与“muapi-drone-style-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/drone-style-video/SKILL.md`
- **muapi-freeze-effect-video**（skill）：Generate a cinematic "freeze effect" video where time stops mid-scene, the subject walks through the frozen world, then time resumes with a snap. 激活提示：当用户任务与“muapi-freeze-effect-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/freeze-effect-video/SKILL.md`
- **muapi-giant-product-showcase**（skill）：Create a dramatic "Giant Product" visual where a regular item is showcased as a massive, building-sized object next to a person, then optionally animate the scene. 激活提示：当用户任务与“muapi-giant-product-showcase”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/giant-product-showcase/SKILL.md`
- **muapi-jewelry-product-video**（skill）：Create a luxury jewelry advertisement with high-end commercial cinematography and detailed macro animation. 激活提示：当用户任务与“muapi-jewelry-product-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/jewelry-product-video/SKILL.md`
- **muapi-music-video**（skill）：Build a short music video from a song theme — N keyframes, animate each, generate matching music. 激活提示：当用户任务与“muapi-music-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/music-video/SKILL.md`
- **muapi-one-shot-video**（skill）：Generate a single continuous cinematic shot video — no cuts, one seamless flowing scene with dramatic lighting and motion. 激活提示：当用户任务与“muapi-one-shot-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/one-shot-video/SKILL.md`
- **muapi-product-ad-cinematic**（skill）：Cinematic 5–10s product ad from a product photo + brand brief. 激活提示：当用户任务与“muapi-product-ad-cinematic”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/product-ad-cinematic/SKILL.md`
- **muapi-product-showcase-video**（skill）：Create a dynamic product showcase with explosive ingredient arrangements, followed by a realistic motion animation. 激活提示：当用户任务与“muapi-product-showcase-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/product-showcase-video/SKILL.md`
- **muapi-product-video-ad-maker**（skill）：Create a high-end cinematic product video advertisement starting from a simple product photo. 激活提示：当用户任务与“muapi-product-video-ad-maker”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/product-video-ad-maker/SKILL.md`
- **muapi-seedance-2**（skill）：Expert Cinema Director skill for Seedance 2.0 ByteDance — high-fidelity video generation across Chinese, Global, and VIP tiers. Supports text-to-video, image-to-video, first-last-frame, omni reference, character training, omni-reference training, video editing, and watermark removal. 激活提示：当用户任务与“muapi-seedance-2”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/seedance-2/SKILL.md`
- **muapi-storyboard-to-cooking-video**（skill）：Turn a single photo of a person into a 15-second cinematic pasta-making or other cuisine tutorial video. First builds a composite reference sheet character + kitchen + 9-step action board , then animates the full cooking sequence with audio in a single continuous shot. 激活提示：当用户任务与“muapi-storyboard-to-cooking-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/storyboard-to-cooking-video/SKILL.md`
- **muapi-talking-baby-video**（skill）：Create a viral-style video of a talking baby with custom costumes and scripts. 激活提示：当用户任务与“muapi-talking-baby-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/talking-baby-video/SKILL.md`
- **muapi-ugc-lifestyle-try-on**（skill）：Generate UGC-style User Generated Content lifestyle photos of a person wearing or using your product — authentic, relatable, social-media-native imagery. 激活提示：当用户任务与“muapi-ugc-lifestyle-try-on”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/ugc-lifestyle-try-on/SKILL.md`
- **muapi-ugc-video-factory**（skill）：Turn a person photo + a product photo + an optional script into a vertical 9:16 UGC-style video ad. Generates a lifestyle hero image Nano-Banana Pro Edit , then animates it with native audio using Seedance 2.0 VIP image-to-video. 激活提示：当用户任务与“muapi-ugc-video-factory”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/motion/ugc-video-factory/SKILL.md`
- **muapi-instagram-post**（skill）：Create a polished, on-brand Instagram post — square or portrait hero image with matching caption and hashtags. 激活提示：当用户任务与“muapi-instagram-post”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/social/instagram-post/SKILL.md`
- **muapi-product-campaign**（skill）：Generate a full multi-channel product campaign — hero visuals, social media assets, short ad video, and platform-specific crops for an end-to-end launch campaign. 激活提示：当用户任务与“muapi-product-campaign”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/social/product-campaign/SKILL.md`
- **muapi-rednote-cover**（skill）：Create a Xiaohongshu RedNote/小红书 style cover image — vibrant, lifestyle-focused, with aesthetic typography overlay suitable for the Chinese social platform. 激活提示：当用户任务与“muapi-rednote-cover”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/social/rednote-cover/SKILL.md`
- **muapi-social-media-video**（skill）：Brand-aware social media video creator. Reads brand-identity.md, ICP.md, and messaging.md to write a post/storyboard, craft an optimized Seedance 2.0 Director prompt, generate reference frames with the best available image model, and produce platform-ready video. 激活提示：当用户任务与“muapi-social-media-video”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/social/social-media-video/SKILL.md`
- **muapi-social-pack**（skill）：Re-render a hero image into Instagram, TikTok, YouTube-shorts and Twitter/X aspect ratios. 激活提示：当用户任务与“muapi-social-pack”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/social/social-pack/SKILL.md`
- **muapi-ugc-ads-workflow**（skill）：Create a User-Generated Content UGC video ad by combining a human selfie and a product image, then generating a video script and an animated ad. 激活提示：当用户任务与“muapi-ugc-ads-workflow”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/social/ugc-ads-workflow/SKILL.md`
- **muapi-youtube-shorts**（skill）：Auto-generate viral 9:16 YouTube Shorts or TikTok / Reels clips from a long-form video. Thin platform-aware wrapper around the AI Clipping skill — picks sensible defaults for short-form social platforms 9:16, 30–60s sweet spot and delegates the actual highlight extraction + crop to muapi.ai's /ai-clipping endpoint. 激活提示：当用户任务与“muapi-youtube-shorts”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/social/youtube-shorts/SKILL.md`
- **muapi-action-figure-generator**（skill）：Convert a photo of a person into a custom 3D action figure, complete with collectible toy packaging. 激活提示：当用户任务与“muapi-action-figure-generator”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/action-figure-generator/SKILL.md`
- **muapi-ad-creative**（skill）：Generate a high-converting ad creative set — hero image, ad copy variations, and platform-optimized crops for Meta, Google Display, and LinkedIn. 激活提示：当用户任务与“muapi-ad-creative”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/ad-creative/SKILL.md`
- **muapi-amazon-product-listing**（skill）：Generate a complete Amazon product listing image set — hero image, lifestyle shot, infographic with features, and comparison/detail closeups optimized for Amazon standards. 激活提示：当用户任务与“muapi-amazon-product-listing”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/amazon-product-listing/SKILL.md`
- **muapi-blog-header**（skill）：Create a professional, eye-catching blog post header image sized for web 1200×628 with optional title composition guidance. 激活提示：当用户任务与“muapi-blog-header”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/blog-header/SKILL.md`
- **muapi-brand-kit**（skill）：Generate a cohesive brand visual kit — logo concept, color palette moodboard, and typography pairing suggestions. 激活提示：当用户任务与“muapi-brand-kit”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/brand-kit/SKILL.md`
- **muapi-brochures**（skill）：Generate a professional multi-page brochure design — cover, inner spread, and back cover — for business, real estate, events, or product launches. 激活提示：当用户任务与“muapi-brochures”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/brochures/SKILL.md`
- **muapi-chibi-collage-effect**（skill）：Turn a real lifestyle photo into a polished "chibi clone sticker diary" image — the original person stays photorealistic, surrounded by 5–8 kawaii chibi mini-clones, scrapbook doodles, and handwritten-style captions that match the scene. 激活提示：当用户任务与“muapi-chibi-collage-effect”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/chibi-collage-effect/SKILL.md`
- **muapi-color-analysis-board**（skill）：Turn a portrait photo into a high-end editorial "Color Analysis Board" in a luxury fashion-magazine style Dior / Ralph Lauren aesthetic — best colors, undertone, makeup guide, capsule wardrobe, hair & jewelry recommendations, all laid out on a clean beige/ivory grid. 激活提示：当用户任务与“muapi-color-analysis-board”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/color-analysis-board/SKILL.md`
- **muapi-couple-grid-creator**（skill）：Create a stylized 6-box grid featuring a couple in various romantic poses and outfits, with each pose framed inside a unique cardboard box packaging. 激活提示：当用户任务与“muapi-couple-grid-creator”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/couple-grid-creator/SKILL.md`
- **muapi-design-guide**（skill）：Create a comprehensive brand design guide — color palette, typography pairings, UI component previews, and visual identity rules with example mockups. 激活提示：当用户任务与“muapi-design-guide”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/design-guide/SKILL.md`
- **muapi-fashion-try-on**（skill）：Virtually try on different outfits by combining a person's photo and a clothing item, then optionally generate a professional fashion model video. 激活提示：当用户任务与“muapi-fashion-try-on”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/fashion-try-on/SKILL.md`
- **muapi-floor-plan-rendering**（skill）：Design a 2D floor plan and convert it into a realistic, high-quality 3D architectural rendering. 激活提示：当用户任务与“muapi-floor-plan-rendering”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/floor-plan-rendering/SKILL.md`
- **muapi-interior-design-visualizer**（skill）：Visualize interior design by generating an empty room and filling it with stylish furniture and decor, or by redesigning an existing room. 激活提示：当用户任务与“muapi-interior-design-visualizer”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/interior-design-visualizer/SKILL.md`
- **muapi-interior-design**（skill）：Create professional interior design visualizations — redesign existing rooms, generate new room concepts, or visualize specific furniture styles in a space. 激活提示：当用户任务与“muapi-interior-design”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/interior-design/SKILL.md`
- **muapi-keyboard-art-maker**（skill）：Generate artistic top-down photos of keyboard keycaps arranged to spell out custom text messages. 激活提示：当用户任务与“muapi-keyboard-art-maker”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/keyboard-art-maker/SKILL.md`
- **muapi-logo-branding**（skill）：Design a professional logo with full branding package — primary logo, variations dark/light/icon-only , color palette, and real-world application mockups. 激活提示：当用户任务与“muapi-logo-branding”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/logo-branding/SKILL.md`
- **muapi-logo-creator**（skill）：Engineer professional-grade brand logos using geometric primitives and negative space — generates minimalist, scalable vector-style marks via muapi.ai 激活提示：当用户任务与“muapi-logo-creator”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/logo-creator/SKILL.md`
- **muapi-logo-generator**（skill）：Quickly generate a single polished logo for any brand — fast, focused, single output with clean vector aesthetic and accurate brand name text. 激活提示：当用户任务与“muapi-logo-generator”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/logo-generator/SKILL.md`
- **muapi-multi-angle-reshoot**（skill）：Re-render a subject or scene from multiple dramatic camera angles, such as fish-eye, bird's-eye, low-angle, and macro, while maintaining consistent identity and detail. 激活提示：当用户任务与“muapi-multi-angle-reshoot”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/multi-angle-reshoot/SKILL.md`
- **muapi-multi-angle-shots**（skill）：Generate a complete set of multi-angle product shots — front, side, back, top-down, and 45-degree perspective — for comprehensive product visualization. 激活提示：当用户任务与“muapi-multi-angle-shots”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/multi-angle-shots/SKILL.md`
- **muapi-nano-banana**（skill）：Reasoning-driven image generation using structured creative briefs Gemini 3 style — generates high-fidelity images via muapi.ai with logic-based prompting 激活提示：当用户任务与“muapi-nano-banana”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/nano-banana/SKILL.md`
- **muapi-photo-pack-generator**（skill）：Generate a pack of professional or aesthetic photos from a single reference image while preserving the exact identity of the person. 激活提示：当用户任务与“muapi-photo-pack-generator”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/photo-pack-generator/SKILL.md`
- **muapi-selfie-with-celebrities**（skill）：Generate a realistic behind-the-scenes selfie of the user with a celebrity or main actor from a specific movie, followed by an option to generate a cinematic long-take video connecting multiple selfies. 激活提示：当用户任务与“muapi-selfie-with-celebrities”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/selfie-with-celebrities/SKILL.md`
- **muapi-storyboard**（skill）：Generate N keyframes for a short story or scene sequence image only, no video . 激活提示：当用户任务与“muapi-storyboard”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/storyboard/SKILL.md`
- **muapi-ui-design**（skill）：Generate high-fidelity UI/UX mockups for mobile and web apps using Atomic Design principles — creates wireframes and design systems via muapi.ai 激活提示：当用户任务与“muapi-ui-design”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/ui-design/SKILL.md`
- **muapi-url-to-design**（skill）：Analyze a website URL and generate a redesigned, improved UI — recreate the visual design with modern aesthetics, better hierarchy, and fresh brand direction. 激活提示：当用户任务与“muapi-url-to-design”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/url-to-design/SKILL.md`
- **muapi-youtube-thumbnail**（skill）：Design a high-CTR YouTube thumbnail — striking imagery, bold text placement, and emotional face/subject if needed. 激活提示：当用户任务与“muapi-youtube-thumbnail”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/visual/youtube-thumbnail/SKILL.md`
- **muapi-workflow**（skill）：Build, run, and visualize multi-step AI generation workflows. The AI architect translates natural language descriptions into connected node graphs — chain image generation, video creation, enhancement, and editing into automated pipelines. 激活提示：当用户任务与“muapi-workflow”描述的流程高度相关时，先用它做安装前体验，再决定是否安装。 证据：`library/workflow/SKILL.md`

## 证据索引

- 共索引 73 条证据。

- **🎭 Generative Media Skills for AI Agents**（documentation）：🎭 Generative Media Skills for AI Agents 证据：`README.md`
- **✏️ MuAPI Media Editing & Enhancement**（skill_instruction）：✏️ MuAPI Media Editing & Enhancement 证据：`core/edit/SKILL.md`
- **🎨 MuAPI Media Generation**（skill_instruction）：Schema-driven generation primitives for images, videos, and audio. 证据：`core/media/SKILL.md`
- **⚙️ MuAPI Platform Utilities**（skill_instruction）：Setup and polling utilities for the muapi.ai platform. 证据：`core/platform/SKILL.md`
- **AI Clipping**（skill_instruction）：One API call: long video in → ranked vertical short clips out. 证据：`library/edit/ai-clipping/SKILL.md`
- **3D Logo Animation**（skill_instruction）：Transform a 2D logo into a premium 3D version and animate it with professional cinematic effects. 证据：`library/motion/3d-logo-animation/SKILL.md`
- **AI Fight Scene Generator**（skill_instruction）：Generate a high-cut-density action / fight scene by first composing a 16-cell storyboard image, then driving Seedance 2.0 image-to-video off that storyboard. 证据：`library/motion/ai-fight-scene/SKILL.md`
- **Animal Vlogger Video**（skill_instruction）：Create a hilarious and ultra-realistic video of an anthropomorphic animal acting like a human vlogger in a real-world setting. 证据：`library/motion/animal-video-generator/SKILL.md`
- **Award Ceremony Video**（skill_instruction）：Generate a 15-second cinematic awards-ceremony video — a host announces a winner from the stage, a spotlight finds them in the crowd, they walk up to the podium, receive the award, and the LED display reveals their name and "THE BEST ACTOR". 证据：`library/motion/award-ceremony-video/SKILL.md`
- **Cartoon Dance Animation**（skill_instruction）：Convert a photo of a person into a Pixar-style 3D cartoon character, then animate it using a reference dance or motion video. 证据：`library/motion/cartoon-dance-animation/SKILL.md`
- **Character Story Video**（skill_instruction）：Create a multi-part animated story video by first establishing a consistent character and then generating sequential scenes and animating them. 证据：`library/motion/character-story-video/SKILL.md`
- **🎬 AI Cinema Director Skill**（skill_instruction）：A specialized skill for AI Agents to direct high-fidelity cinematic video. The Cinema Director skill translates high-level creative intent into technical cinematographic directives for state-of-the-art video models Veo3, Kling, Luma . 证据：`library/motion/cinema-director/SKILL.md`
- **Drone-Style Video**（skill_instruction）：Generate aerial drone-perspective footage — sweeping bird's-eye views, orbit shots, and flyover sequences for landscapes, architecture, and events. 证据：`library/motion/drone-style-video/SKILL.md`
- **Freeze Effect Video**（skill_instruction）：Generate a cinematic "freeze effect" video where time stops mid-scene, the subject walks through the frozen world, then time resumes with a snap. 证据：`library/motion/freeze-effect-video/SKILL.md`
- **Giant Product Showcase**（skill_instruction）：Create a dramatic "Giant Product" visual where a regular item is showcased as a massive, building-sized object next to a person, then optionally animate the scene. 证据：`library/motion/giant-product-showcase/SKILL.md`
- **Jewelry Product Video**（skill_instruction）：Create a luxury jewelry advertisement with high-end commercial cinematography and detailed macro animation. 证据：`library/motion/jewelry-product-video/SKILL.md`
- **Music Video**（skill_instruction）：Build a short music video from a song theme — N keyframes, animate each, generate matching music. 证据：`library/motion/music-video/SKILL.md`
- **One-Shot Video**（skill_instruction）：Generate a single continuous cinematic shot video — no cuts, one seamless flowing scene with dramatic lighting and motion. 证据：`library/motion/one-shot-video/SKILL.md`
- **Cinematic Product Ad**（skill_instruction）：Cinematic 5–10s product ad from a product photo + brand brief. 证据：`library/motion/product-ad-cinematic/SKILL.md`
- **Product Showcase Video**（skill_instruction）：Create a dynamic product showcase with explosive ingredient arrangements, followed by a realistic motion animation. 证据：`library/motion/product-showcase-video/SKILL.md`
- **Product Video Ad Maker**（skill_instruction）：Create a high-end cinematic product video advertisement starting from a simple product photo. 证据：`library/motion/product-video-ad-maker/SKILL.md`
- **🎬 Seedance 2.0 Cinema Expert**（skill_instruction）：The definitive skill for "Director-Level" AI video orchestration. Seedance 2.0 is not a descriptive model; it is an instructional model. It responds best to technical cinematography, physics directives, and precise camera grammar. 证据：`library/motion/seedance-2/SKILL.md`
- **Storyboard to Cooking Video**（skill_instruction）：Turn a single photo of a person into a polished 15-second cinematic cooking tutorial. The skill first generates a high-end production reference sheet — character look, kitchen environment, and a 9-panel action board — then drives a continuous reference-to-video render that keeps the subject's face, outfit, and kitchen consistent across every frame. 证据：`library/motion/storyboard-to-cooking-video/SKILL.md`
- **Talking Baby Video**（skill_instruction）：Create a viral-style video of a talking baby with custom costumes and scripts. 证据：`library/motion/talking-baby-video/SKILL.md`
- **UGC Lifestyle Try-On**（skill_instruction）：Generate UGC-style User Generated Content lifestyle photos of a person wearing or using your product — authentic, relatable, social-media-native imagery. 证据：`library/motion/ugc-lifestyle-try-on/SKILL.md`
- **UGC Video Factory**（skill_instruction）：Turn a person photo + product photo + optional script & environment into a vertical 9:16 UGC-style video ad with native dialogue audio. 证据：`library/motion/ugc-video-factory/SKILL.md`
- **Instagram Post**（skill_instruction）：Create a polished, on-brand Instagram post — square or portrait hero image with matching caption and hashtags. 证据：`library/social/instagram-post/SKILL.md`
- **Product Campaign Pack**（skill_instruction）：Generate a full multi-channel product campaign — hero visuals, social media assets, short ad video, and platform-specific crops for an end-to-end launch campaign. 证据：`library/social/product-campaign/SKILL.md`
- **RedNote Cover**（skill_instruction）：Create a Xiaohongshu RedNote/小红书 style cover image — vibrant, lifestyle-focused, with aesthetic typography overlay suitable for the Chinese social platform. 证据：`library/social/rednote-cover/SKILL.md`
- **Social Media Video Creator**（skill_instruction）：End-to-end pipeline: Brand Files → Storyboard → Reference Images → Seedance 2.0 Video. 证据：`library/social/social-media-video/SKILL.md`
- **Social Media Pack**（skill_instruction）：Re-render a hero image into Instagram, TikTok, YouTube-shorts and Twitter/X aspect ratios. 证据：`library/social/social-pack/SKILL.md`
- **UGC Ads Workflow**（skill_instruction）：Create a User-Generated Content UGC video ad by combining a human selfie and a product image, then generating a video script and an animated ad. 证据：`library/social/ugc-ads-workflow/SKILL.md`
- **YouTube Shorts Generator**（skill_instruction）：Long video → ranked vertical short clips, tuned for short-form social. 证据：`library/social/youtube-shorts/SKILL.md`
- **Action Figure Generator**（skill_instruction）：Convert a photo of a person into a custom 3D action figure, complete with collectible toy packaging. 证据：`library/visual/action-figure-generator/SKILL.md`
- **Ad Creative Set**（skill_instruction）：Generate a high-converting ad creative set — hero image, ad copy variations, and platform-optimized crops for Meta, Google Display, and LinkedIn. 证据：`library/visual/ad-creative/SKILL.md`
- **Amazon Product Listing Pack**（skill_instruction）：Generate a complete Amazon product listing image set — hero image, lifestyle shot, infographic with features, and comparison/detail closeups optimized for Amazon standards. 证据：`library/visual/amazon-product-listing/SKILL.md`
- **Blog Header**（skill_instruction）：Create a professional, eye-catching blog post header image sized for web 1200×628 with optional title composition guidance. 证据：`library/visual/blog-header/SKILL.md`
- **Brand Kit**（skill_instruction）：Generate a cohesive brand visual kit — logo concept, color palette moodboard, and typography pairing suggestions. 证据：`library/visual/brand-kit/SKILL.md`
- **Brochure Designer**（skill_instruction）：Generate a professional multi-page brochure design — cover, inner spread, and back cover — for business, real estate, events, or product launches. 证据：`library/visual/brochures/SKILL.md`
- **Chibi Collage Effect**（skill_instruction）：Turn a real lifestyle photo into a polished "chibi clone sticker diary" image — the original person stays photorealistic, surrounded by 5–8 kawaii chibi mini-clones, scrapbook doodles, and handwritten-style captions that match the scene. 证据：`library/visual/chibi-collage-effect/SKILL.md`
- **Color Analysis Board**（skill_instruction）：Turn a portrait photo into a high-end editorial "Color Analysis Board" in a luxury fashion-magazine style Dior / Ralph Lauren aesthetic — best colors, undertone, makeup guide, capsule wardrobe, hair & jewelry recommendations, all laid out on a clean beige/ivory grid. 证据：`library/visual/color-analysis-board/SKILL.md`
- **Couple Grid Creator**（skill_instruction）：Create a stylized 6-box grid featuring a couple in various romantic poses and outfits, with each pose framed inside a unique cardboard box packaging. 证据：`library/visual/couple-grid-creator/SKILL.md`
- **Brand Design Guide**（skill_instruction）：Create a comprehensive brand design guide — color palette, typography pairings, UI component previews, and visual identity rules with example mockups. 证据：`library/visual/design-guide/SKILL.md`
- **Fashion Try-On**（skill_instruction）：Virtually try on different outfits by combining a person's photo and a clothing item, then optionally generate a professional fashion model video. 证据：`library/visual/fashion-try-on/SKILL.md`
- **Floor Plan Rendering**（skill_instruction）：Design a 2D floor plan and convert it into a realistic, high-quality 3D architectural rendering. 证据：`library/visual/floor-plan-rendering/SKILL.md`
- **Interior Design Visualizer**（skill_instruction）：Visualize interior design by generating an empty room and filling it with stylish furniture and decor, or by redesigning an existing room. 证据：`library/visual/interior-design-visualizer/SKILL.md`
- **Interior Design**（skill_instruction）：Create professional interior design visualizations — redesign existing rooms, generate new room concepts, or visualize specific furniture styles in a space. 证据：`library/visual/interior-design/SKILL.md`
- **Keyboard Art Maker**（skill_instruction）：Generate artistic top-down photos of keyboard keycaps arranged to spell out custom text messages. 证据：`library/visual/keyboard-art-maker/SKILL.md`
- **Logo + Branding Package**（skill_instruction）：Design a professional logo with full branding package — primary logo, variations dark/light/icon-only , color palette, and real-world application mockups. 证据：`library/visual/logo-branding/SKILL.md`
- **🖼️ Logo Creator Skill**（skill_instruction）：A specialized skill for AI Agents to engineer professional-grade brand identities. The Logo Creator skill translates brand vision into minimalist, scalable, and iconic vector-style marks. 证据：`library/visual/logo-creator/SKILL.md`
- **Logo Generator**（skill_instruction）：Quickly generate a single polished logo for any brand — fast, focused, single output with clean vector aesthetic and accurate brand name text. 证据：`library/visual/logo-generator/SKILL.md`
- **Multi-Angle Reshoot**（skill_instruction）：Re-render a subject or scene from multiple dramatic camera angles, such as fish-eye, bird's-eye, low-angle, and macro, while maintaining consistent identity and detail. 证据：`library/visual/multi-angle-reshoot/SKILL.md`
- **Multi-Angle Shots**（skill_instruction）：Generate a complete set of multi-angle product shots — front, side, back, top-down, and 45-degree perspective — for comprehensive product visualization. 证据：`library/visual/multi-angle-shots/SKILL.md`
- **🍌 Nano-Banana Expert Skill Gemini 3 Style**（skill_instruction）：🍌 Nano-Banana Expert Skill Gemini 3 Style 证据：`library/visual/nano-banana/SKILL.md`
- **📸 Photo Pack Generator Expert Skill Identity-Lock Edition**（skill_instruction）：📸 Photo Pack Generator Expert Skill Identity-Lock Edition 证据：`library/visual/photo-pack-generator/SKILL.md`
- **Selfie with Celebrities**（skill_instruction）：Generate a realistic behind-the-scenes selfie of the user with a celebrity or main actor from a specific movie, followed by an option to generate a cinematic long-take video connecting multiple selfies. 证据：`library/visual/selfie-with-celebrities/SKILL.md`
- **Storyboard Generator**（skill_instruction）：Generate N keyframes for a short story or scene sequence image only, no video . 证据：`library/visual/storyboard/SKILL.md`
- **🎨 UI/UX Design Mockup Skill**（skill_instruction）：A specialized skill for AI Agents to architect high-fidelity digital interfaces. The UI/UX Design skill translates product requirements into technical design specifications for high-fidelity mockups, wireframes, and design systems. 证据：`library/visual/ui-design/SKILL.md`
- **URL to Design**（skill_instruction）：Analyze a website URL and generate a redesigned, improved UI — recreate the visual design with modern aesthetics, better hierarchy, and fresh brand direction. 证据：`library/visual/url-to-design/SKILL.md`
- **YouTube Thumbnail**（skill_instruction）：Design a high-CTR YouTube thumbnail — striking imagery, bold text placement, and emotional face/subject if needed. 证据：`library/visual/youtube-thumbnail/SKILL.md`
- 其余 13 条证据见 `AI_CONTEXT_PACK.json` 或 `EVIDENCE_INDEX.json`。

## 宿主 AI 必须遵守的规则

- **把本资产当作开工前上下文，而不是运行环境。**：AI Context Pack 只包含证据化项目理解，不包含目标项目的可执行状态。 证据：`README.md`, `core/edit/SKILL.md`, `core/media/SKILL.md`
- **回答用户时区分可预览内容与必须安装后才能验证的内容。**：安装前体验的消费者价值来自降低误装和误判，而不是伪装成真实运行。 证据：`README.md`, `core/edit/SKILL.md`, `core/media/SKILL.md`

## 用户开工前应该回答的问题

- 你准备在哪个宿主 AI 或本地环境中使用它？
- 你只是想先体验工作流，还是准备真实安装？
- 你最在意的是安装成本、输出质量、还是和现有规则的冲突？

## 验收标准

- 所有能力声明都能回指到 evidence_refs 中的文件路径。
- AI_CONTEXT_PACK.md 没有把预览包装成真实运行。
- 用户能在 3 分钟内看懂适合谁、能做什么、如何开始和风险边界。

---

## Doramagic Context Augmentation

下面内容用于强化 Repomix/AI Context Pack 主体。Human Manual 只提供阅读骨架；踩坑日志会被转成宿主 AI 必须遵守的工作约束。

## Human Manual 骨架

使用规则：这里只是项目阅读路线和显著性信号，不是事实权威。具体事实仍必须回到 repo evidence / Claim Graph。

宿主 AI 硬性规则：
- 不得把页标题、章节顺序、摘要或 importance 当作项目事实证据。
- 解释 Human Manual 骨架时，必须明确说它只是阅读路线/显著性信号。
- 能力、安装、兼容性、运行状态和风险判断必须引用 repo evidence、source path 或 Claim Graph。

- **项目介绍**：importance `high`
  - source_paths: README.md
- **快速入门指南**：importance `high`
  - source_paths: core/platform/setup.sh, README.md
- **系统架构设计**：importance `high`
  - source_paths: README.md, core/edit/SKILL.md, core/platform/SKILL.md
- **Schema数据参考**：importance `medium`
  - source_paths: schema_data.json
- **核心原语层详解**：importance `high`
  - source_paths: core/edit/SKILL.md, core/edit/edit-image.sh, core/edit/enhance-image.sh, core/edit/lipsync.sh, core/edit/video-effects.sh
- **视频与动效技能库**：importance `high`
  - source_paths: library/motion/cinema-director/SKILL.md, library/motion/seedance-2/SKILL.md, library/motion/seedance-2/scripts/generate-seedance.sh, library/motion/cinema-director/scripts/generate-film.sh, library/motion/product-video-ad-maker/SKILL.md
- **图像与设计技能库**：importance `high`
  - source_paths: library/visual/nano-banana/SKILL.md, library/visual/nano-banana/scripts/generate-nano-art.sh, library/visual/ui-design/SKILL.md, library/visual/ui-design/scripts/generate-mockup.sh, library/visual/logo-creator/SKILL.md
- **社交媒体技能库**：importance `medium`
  - source_paths: library/social/youtube-shorts/SKILL.md, library/social/youtube-shorts/scripts/run-youtube-shorts.sh, library/social/instagram-post/SKILL.md, library/social/product-campaign/SKILL.md, library/social/rednote-cover/SKILL.md

## Repo Inspection Evidence / 源码检查证据

- repo_clone_verified: true
- repo_inspection_verified: true
- repo_commit: `31c49f1b01aa54a809454aadfc35a5e817f17a2e`
- inspected_files: `README.md`

宿主 AI 硬性规则：
- 没有 repo_clone_verified=true 时，不得声称已经读过源码。
- 没有 repo_inspection_verified=true 时，不得把 README/docs/package 文件判断写成事实。
- 没有 quick_start_verified=true 时，不得声称 Quick Start 已跑通。

## Doramagic Pitfall Constraints / 踩坑约束

这些规则来自 Doramagic 发现、验证或编译过程中的项目专属坑点。宿主 AI 必须把它们当作工作约束，而不是普通说明文字。

### Constraint 1: 来源证据：Optional: add a 'publish to Vynly' skill after media generation?

- Trigger: GitHub 社区证据显示该项目存在一个安全/权限相关的待验证问题：Optional: add a 'publish to Vynly' skill after media generation?
- Host AI rule: 来源问题仍为 open，Pack Agent 需要复核是否仍影响当前版本。
- Why it matters: 可能影响授权、密钥配置或安全边界。
- Evidence: community_evidence:github | cevd_e801ed325bcf4fbbb8e0d9cac02b5f7f | https://github.com/SamurAIGPT/Generative-Media-Skills/issues/89 | 来源类型 github_issue 暴露的待验证使用条件。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 2: 仓库名和安装名不一致

- Trigger: 仓库名 `generative-media-skills` 与安装入口 `skills` 不完全一致。
- Host AI rule: 在 npm/PyPI/GitHub 上确认包名映射和官方 README 说明。
- Why it matters: 用户照着仓库名搜索包或照着包名找仓库时容易走错入口。
- Evidence: identity.distribution | github_repo:645381450 | https://github.com/SamurAIGPT/Generative-Media-Skills | repo=generative-media-skills; install=skills
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 3: 来源证据：Error downloading model

- Trigger: GitHub 社区证据显示该项目存在一个安装相关的待验证问题：Error downloading model
- Host AI rule: 来源显示可能已有修复、规避或版本变化，说明书中必须标注适用版本。
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | cevd_ec1cdc92d8c84b2bbce43cf37a668443 | https://github.com/SamurAIGPT/Generative-Media-Skills/issues/44 | 来源讨论提到 node 相关条件，需在安装/试用前复核。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 4: 来源证据：Server install pre-requisites on Ubuntu

- Trigger: GitHub 社区证据显示该项目存在一个安装相关的待验证问题：Server install pre-requisites on Ubuntu
- Host AI rule: 来源显示可能已有修复、规避或版本变化，说明书中必须标注适用版本。
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | cevd_c7a0e07d61b547aba3280ac82ac305e2 | https://github.com/SamurAIGPT/Generative-Media-Skills/issues/43 | 来源讨论提到 python 相关条件，需在安装/试用前复核。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 5: 来源证据：install vicuna13b?

- Trigger: GitHub 社区证据显示该项目存在一个安装相关的待验证问题：install vicuna13b?
- Host AI rule: 来源显示可能已有修复、规避或版本变化，说明书中必须标注适用版本。
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | cevd_c6ace8e72f95491f945200849e083d96 | https://github.com/SamurAIGPT/Generative-Media-Skills/issues/46 | 来源类型 github_issue 暴露的待验证使用条件。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 6: 来源证据：npm run dev error

- Trigger: GitHub 社区证据显示该项目存在一个安装相关的待验证问题：npm run dev error
- Host AI rule: 来源显示可能已有修复、规避或版本变化，说明书中必须标注适用版本。
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | cevd_de5cf59443c74838a8d10d1ecbec9457 | https://github.com/SamurAIGPT/Generative-Media-Skills/issues/45 | 来源讨论提到 python 相关条件，需在安装/试用前复核。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 7: 来源证据：npm run dev hang in certain point

- Trigger: GitHub 社区证据显示该项目存在一个安装相关的待验证问题：npm run dev hang in certain point
- Host AI rule: 来源显示可能已有修复、规避或版本变化，说明书中必须标注适用版本。
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | cevd_8082f7c5e7be496c89fac789d932e74c | https://github.com/SamurAIGPT/Generative-Media-Skills/issues/34 | 来源讨论提到 node 相关条件，需在安装/试用前复核。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 8: 可能修改宿主 AI 配置

- Trigger: 项目面向 Claude/Cursor/Codex/Gemini/OpenCode 等宿主，或安装命令涉及用户配置目录。
- Host AI rule: 列出会写入的配置文件、目录和卸载/回滚步骤。
- Why it matters: 安装可能改变本机 AI 工具行为，用户需要知道写入位置和回滚方法。
- Evidence: capability.host_targets | github_repo:645381450 | https://github.com/SamurAIGPT/Generative-Media-Skills | host_targets=claude, claude_code, cursor
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 9: 来源证据：Model download error on 100% progress

- Trigger: GitHub 社区证据显示该项目存在一个配置相关的待验证问题：Model download error on 100% progress
- Host AI rule: 来源显示可能已有修复、规避或版本变化，说明书中必须标注适用版本。
- Why it matters: 可能增加新用户试用和生产接入成本。
- Evidence: community_evidence:github | cevd_14498a89412f40ceb03d61889ea96de2 | https://github.com/SamurAIGPT/Generative-Media-Skills/issues/38 | 来源讨论提到 python 相关条件，需在安装/试用前复核。
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 10: 能力判断依赖假设

- Trigger: README/documentation is current enough for a first validation pass.
- Host AI rule: 将假设转成下游验证清单。
- Why it matters: 假设不成立时，用户拿不到承诺的能力。
- Evidence: capability.assumptions | github_repo:645381450 | https://github.com/SamurAIGPT/Generative-Media-Skills | README/documentation is current enough for a first validation pass.
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。