digital-human-generator

name	digital-human-generator
description	NexTide 数字人生成。使用人物图片/视频、音频和口播文本创建图片源或视频源数字人口播任务。适合“生成数字人视频”“用这张人物图做口播”“图片数字人/视频数字人”等请求。
allowed-tools	Read, Write, Bash

Digital Human Generator

Follow shared NexTide rules in:

• nextide-shared

Use this skill when the user wants to create a digital human video from:

• person image + audio
• person image + audio + script
• person video + audio

Capability

Capability id:

digital-human.video.generate

CLI contract:

npm run nextide -- capability run digital-human.video.generate \
  --input .nextide/input/digital-human-video.json \
  --output .nextide/output/digital-human-video-result.json \
  --mode submit \
  --user-api-key <NEX用户积分API_KEY>

Input JSON

Image source lip-sync:

{
  "sourceType": "IMAGE",
  "imageUrl": "https://.../person.png",
  "audioUrl": "https://.../voice.mp3",
  "type": "LIP_SYNC"
}

Voice clone / script-aware task:

{
  "sourceType": "IMAGE",
  "personImage": "https://.../person.png",
  "audioUrl": "https://.../voice.mp3",
  "script": "大家好，今天讲一个...",
  "type": "VOICE_CLONE",
  "durationSeconds": 30
}

Aliases:

• personImage / sourceImage → imageUrl
• personVideo / sourceVideo → videoUrl
• voiceUrl → audioUrl
• script / text → scriptContent

Long-running Rule

Digital human generation can take up to 60 minutes.

The capability usually returns:

waiting_callback

The result contains a data.id, which is the NexTide digitalHumanVideo record id.

Progress can be inspected in NexTide UI or through the underlying API:

GET /api/digital-human/videos/<id>

Output Handling

Expected result shape:

{
  "run": {
    "status": "waiting_callback",
    "result": {
      "data": {
        "id": "...",
        "status": "GENERATING",
        "resultUrl": null
      }
    }
  }
}

If resultUrl is present, export multimodal artifacts and return the local video/preview paths:

RUN_ID=$(node -e "const r=require('./.nextide/output/digital-human-video-result.json'); console.log(r.run && r.run.runId)")
npm run nextide -- run artifacts "$RUN_ID" \
  --output-dir .nextide/output/$RUN_ID \
  --download \
  --gallery

Return:

• generated video URL
• local downloaded video path if available
• preview.html path if generated
• status/task id

Rules

• Do not wait 60 minutes by default.
• Do not claim the video is finished until resultUrl exists or status is completed.
• User must have rights to the face/image/video/audio source.
• This skill does not generate TTS/audio in MVP; require audioUrl.

NexTide Capability Contract

视频数字人生成

• Capability: digital-human.video.generate
• Version: 0.2.0
• Category: video
• Status: available
• Execution: internal_api
• Async: true
• Cost level: high
• Required auth: nexTideApiKey
• Required env: N8N_DIGITAL_HUMAN_WEBHOOK or DIGITAL_HUMAN_WEBHOOK_URL
• Required plan: paid
• Rate limit: 5/minute, 20/hour
• Estimated credits: 30
• Estimated duration: 1200 seconds
• Tags: digital-human, video, long-running

Description:

使用人物图、文案和声音生成口播数字人视频。最长可能 60 分钟。

Examples:

• 图片数字人口播

  {
    "personImage": "https://example.com/person.png",
    "audioUrl": "https://example.com/audio.mp3",
    "script": "口播文案"
  }
  

Input fields:

• personImage (string, required)：人物图片 URL。
• script (string, required)：口播文案。
• voiceId (string)：声音 ID 或预设。
• duration (number)：目标时长，秒。

Output fields:

• videoUrl (string)：生成视频 URL。
• taskId (string)：内部任务 ID。

CLI:

nextide capability run digital-human.video.generate \
  --input .nextide/input/digital-human.video.generate.json \
  --output .nextide/output/digital-human.video.generate-result.json \
  --mode submit \
  --wait \
  --timeout 3600 \
  --interval 5

RUN_ID=$(node -e "const r=require('./.nextide/output/digital-human.video.generate-result.json'); console.log(r.run && r.run.runId)")
nextide run follow "$RUN_ID" \
  --output-dir .nextide/output/$RUN_ID \
  --timeout 3600 \
  --interval 5

Artifact-first reading order:

1. Read .nextide/output/$RUN_ID/summary.json.
2. Read .nextide/output/$RUN_ID/manifest.json.
3. Return preview.html / gallery.html with rich preview when supported.
4. Return datatable.json for data/table results.
5. Return local artifact paths when present.
6. If a remote URL artifact is present, return the URL from manifest.
7. Only inspect the full result JSON when manifest is insufficient.

General Rules

• Use NexTide capability IDs, not internal n8n webhook URLs.
• Do not expose API secrets or internal webhook URLs in prompts or output.
• If status is not available, fail fast and explain what is missing.
• For async tasks, prefer --wait when the user wants a finished result in the same turn.
• After a finished run, use nextide run artifacts <run-id> --output-dir .nextide/output/<run-id> --download --gallery --datatable and read summary.json then manifest.json.
• For long-running runs, prefer nextide run follow <run-id> --output-dir .nextide/output/<run-id> --timeout 1800 --interval 5.
• Prefer returning summary.recommendedResponse.message, preview.html, datatable.json, and local artifact paths over pasting huge raw JSON.
• If the CLI output includes explanation, convert it into a clear user-facing failure message with next actions.