music video mv 制作

Generate MV (Music Video) with AI-driven storyboarding, image generation, and video synthesis.

Security Scan

The skill's requirements and runtime instructions are consistent with a cloud MV-generation service: it needs an OHEYSAI_API_KEY and will upload the user's audio to ohyesai.com to generate videos.

Purpose & Capability

Name/description (AI-driven MV generation) align with required resources: an API key for ohyesai.com and network-capable tools (curl) to upload audio and poll tasks. Required binaries (curl, sleep) and the single env var OHEYSAI_API_KEY are proportionate to the stated purpose.

Instruction Scope

SKILL.md instructs the agent to read the user's audio file, upload it as multipart/form-data to ohyesai.com, submit a task and poll for completion. This stays within the MV-generation purpose, but it does transmit user audio (and uses the API key) to a third-party service — users should be aware of that transfer and any privacy/terms implications.

Install Mechanism

No install spec and no code files (instruction-only) — minimal surface written to disk. This is low-risk and appropriate for a wrapper that delegates work to a web API.

Credentials

Only one required environment variable (OHEYSAI_API_KEY) is declared and used in the documented API calls. The requested secret is proportional to a hosted API service. No unrelated credentials or config paths are requested.

Persistence & Privilege

always is false and the skill has no install-time persistence or system-wide config modifications. It does not request elevated/system privileges or modify other skills.

Assessment

This skill will upload any audio you provide to https://ohyesai.com using the OHEYSAI_API_KEY you supply. Before installing or using it: - Confirm you trust ohyesai.com (review their privacy policy and terms) because your audio will be sent off-device. - Avoid uploading sensitive or confidential audio. - Be aware the API key is passed in the documented endpoints (appears as a query parameter), which can expose keys in logs or referrers — treat the key as sensitive and rotate it if compromised. - The skill requires curl and sleep; no local code is installed. - If you need stronger privacy, prefer on-device tools or a service with clear retention/deletion guarantees.

Skill: 音乐视频 (MV) 生成器 (MV Generator)

1. 技能描述

该技能允许用户上传音频并将其转化为具有特定视觉风格（如动漫、写实）和指定比例（横屏/竖屏）的音乐视频（MV）。

2. 预置条件与配置

• API Key 获取: 引导用户前往 https://ohyesai.com 获取。
• 环境变量: 必须配置 OHEYSAI_API_KEY。
• 文件限制: 音频文件大小不得超过 50MB。

3. 核心参数

在启动流程前，需确保获得以下参数：

• audioFile: 用户上传的音频原始文件。
• style: 视频的视觉风格（如：动漫风格、赛博朋克、唯美古风等）。
• aspectRatio: 画面比例，仅限以下两个值：
  • LANDSCAPE: 横屏 (16:9)
  • PORTRAIT: 竖屏 (9:16)

---

4. 接口定义 (API Reference)

4.1 上传音频接口 (Upload)

• URL: POST https://ohyesai.com/ohyesai-next/api/vio/skill/upload?apikey={{OHEYSAI_API_KEY}}
• Payload: multipart/form-data (字段名: file)
• 功能: 将本地音频转换为 audioFileId。

4.2 提交任务接口 (Submit)

• URL: POST https://ohyesai.com/ohyesai-next/api/vio/skill/mv-submit?apikey={{OHEYSAI_API_KEY}}
• Request Body:

  {
    "audioFileId": "string",
    "style": "string",
    "aspectRatio": "LANDSCAPE | PORTRAIT"
  }
  

• Response: 返回 taskId (String)。

4.3 查询任务接口 (Query)

• URL: GET https://ohyesai.com/ohyesai-next/api/vio/skill/mv-query?apikey={{OHEYSAI_API_KEY}}&taskId={{taskId}}
• Response Structure:

  {
    "code": 0,
    "data": {
      "state": "SUCCESS | RUNNING | FAIL",
      "data": "视频下载URL(仅SUCCESS时存在)"
    }
  }
  

---

5. 执行流程 (Workflow)

第一阶段：准备与校验

1. 文件校验: 接收用户音频文件，检查大小。若超过 50MB，提示用户压缩后重新上传。
2. 获取 FileId: 调用 上传音频接口。如果失败，告知用户服务器连接异常。

第二阶段：参数确认

1. 如果用户未指定 style，主动询问：“您希望视频是什么风格？（例如：动漫、写实、3D、水墨等）”。
2. 如果用户未指定 aspectRatio，询问：“您需要横屏(LANDSCAPE)还是竖屏(PORTRAIT)？”

第三阶段：提交与轮询

1. 提交: 调用 提交任务接口，记录返回的 taskId。
2. 轮询: 初始等待 10 秒，随后每隔 5-10 秒调用一次 查询任务接口。

• RUNNING: 继续等待，并向用户发送进度反馈（如：“正在为您精心制作视频，请稍候...”）。
• FAIL: 停止轮询，告知用户任务失败，并尝试分析原因。
• SUCCESS: 提取 data.data 中的视频链接，并以预览卡片或链接形式发送给用户。

---

6. 异常与边界处理

• API Key 缺失: 若环境变量未配置，引导用户前往官网。
• 格式错误: 若上传非音频文件，拦截并提示。
• 超时处理: 若轮询超过 10 分钟仍为 RUNNING，提示用户记录 taskId 稍后手动查询。

7. 交互示例

User: 帮我把这首歌做成一个动漫风格的竖屏视频。 [附件: song.mp3]

Agent:

1. (Internal) 检查 song.mp3 大小 (<50MB)。
2. (Internal) 调用 /upload 获取 audioFileId。
3. (Internal) 调用 /mv-submit 发起任务，获得 taskId: "f7fb7..."。
4. (Output) 已收到您的请求！正在为您生成动漫风格的竖屏 MV，这可能需要一点时间，我会在这里通知您进度。
5. (Internal/Loop) 调用 /mv-query，状态为 RUNNING。
6. (Internal/Loop) 调用 /mv-query，状态变为 SUCCESS。
7. (Output) 视频制作完成！您可以点击下方链接查看或下载：[视频链接]

联系客服