Chanjing Avatar (Lip-Syncing)#

功能说明#

调用蝉镜 Avatar Open API：上传音视频素材、创建对口型任务、轮询与获取结果链接。脚本为 Python HTTP/上传客户端， 不 依赖 ffmpeg/ffprobe。

运行依赖#

• python3 与同仓库 scripts/*.py
• 无 ffmpeg/ffprobe 门控

环境变量与机器可读声明#

• 环境变量键名与说明： manifest.yaml（ environment 段）及本文
• 变量、凭据、合规 permissions、 clientPermissions、 agentPolicy： manifest.yaml

使用命令#

• ClawHub（slug 以注册表为准）： clawhub run chanjing-avatar
• 本仓库： python skills/chanjing-avatar/scripts/create_task.py …（流程见正文 How to Use）

---

登记与审稿（单一事实来源）#

主凭据、上传/下载边界、浏览器引导等： 以 manifest.yaml 为准。本篇 How to Use 起为 API 步骤说明。

When to Use This Skill#

Use this skill when the user needs to create lip-syncing videos (digital avatar videos) with synchronized mouth movements.

Chanjing Avatar supports:

• Text-driven or audio-driven lip-syncing
• Multiple system voices for TTS
• Video resolution customization
• Task status polling and callback

How to Use This Skill#

前置条件：执行本 Skill 前，必须先通过 chanjing-credentials-guard 完成 AK/SK 与 Token 校验。凭据与审稿对表见 manifest.yaml。

Security & credentials（引用）#

详见 manifest.yaml 中 credentials 与 clientPermissions（含本地上传、结果 URL、浏览器行为；合规见顶层 permissions）。

Multiple APIs need to be invoked. All share the domain: " https://open-api.chanjing.cc". All requests communicate using json. You should use utf-8 to encode and decode text throughout this task.

1. Obtain an access_token, which is required for all subsequent API calls
2. Upload your video/audio files using the File Management API to get file_id
3. Create a lip-syncing task with video and audio/text using these file_id values
4. Poll the Query Task Detail API or use Task List API to check status
5. Download the generated video using the url in response when status is completed

Obtain AccessToken#

从 ~/.chanjing/credentials.json 读取 app_id 和 secret_key，若无有效 Token 则调用：

http

POST /open/v1/access_token
Content-Type: application/json

请求体（使用本地配置的 app_id、secret_key）：

json

{
  "app_id": "<从 credentials.json 读取>",
  "secret_key": "<从 credentials.json 读取>"
}

Response example:

json

{
  "trace_id": "8ff3fcd57b33566048ef28568c6cee96",
  "code": 0,
  "msg": "success",
  "data": {
    "access_token": "1208CuZcV1Vlzj8MxqbO0kd1Wcl4yxwoHl6pYIzvAGoP3DpwmCCa73zmgR5NCrNu",
    "expire_in": 1721289220
  }
}

Response field description:

First-level FieldSecond-level FieldDescriptioncodeResponse status codemsgResponse messagedataResponse dataaccess_tokenValid for one day, previous token will be invalidatedexpire_inToken expiration time

Response Status Code Description

CodeDescription0Success400Invalid parameter format40000Parameter error50000System internal error

Upload Media Files (File Management)#

Before creating a lip-syncing task, you must upload your video (and optional audio) files using the File Management API to obtain file_id values.

The full documentation is here: [File Management](https://doc.chanjing.cc/api/file/file-management.html).

Step 1: Get upload URL

http

GET /open/v1/common/create_upload_url
access_token: {{access_token}}

Query params:

KeyExampleDescriptionservicelip_sync_video / lip_sync_audioFile usage purpose. Use lip_sync_video for driving video, lip_sync_audio for audio (if audio-driven).name1.mp4Original file name including extension

You will get a response containing sign_url, mime_type, and file_id. Use the sign_url with HTTP PUT to upload the file, setting Content-Type to the returned mime_type. After the PUT completes, poll the file detail API until the file is ready (do not assume a fixed wait). Keep the returned file_id for video_file_id / audio_file_id below.

Polling: Call GET /open/v1/common/file_detail?id={{file_id}} with access_token until the response data.status indicates success (e.g. status === 2). Only then use the file_id for the create task API.

Create Lip-Syncing Task#

Submit a lip-syncing video creation task, which returns a video ID for polling later.

http

POST /open/v1/video_lip_sync/create
access_token: {{access_token}}
Content-Type: application/json

Request body example (TTS-driven):

json

{
  "video_file_id": "e284db4d95de4220afe78132158156b5",
  "screen_width": 1080,
  "screen_height": 1920,
  "callback": "https://example.com/openapi/callback",
  "model": 0,
  "audio_type": "tts",
  "tts_config": {
    "text": "君不见黄河之水天上来，奔流到海不复回。",
    "audio_man_id": "C-f2429d07554749839849497589199916",
    "speed": 1,
    "pitch": 1
  }
}

Request body example (Audio-driven):

json

{
  "video_file_id": "e284db4d95de4220afe78132158156b5",
  "screen_width": 1080,
  "screen_height": 1920,
  "model": 0,
  "audio_type": "audio",
  "audio_file_id": "audio_file_id_from_file_management"
}

Request field description:

Parameter NameTypeRequiredDescriptionvideo_file_idstringYesVideo file ID from File Management ( data.file_id). Supports mp4, mov, webmscreen_widthintNoScreen width, default 1080screen_heightintNoScreen height, default 1920backwayintNoPlayback order when reaching end: 1-normal, 2-reverse. Default 1drive_modestringNoDrive mode: ""-normal, "random"-random frame. Default ""callbackstringNoCallback URL for async notificationmodelintNoModel version: 0-basic, 1-high quality. Default 0audio_typestringNoAudio type: "tts"-text driven, "audio"-audio driven. Default "tts"tts_configobjectYes (for tts)TTS configuration when audio_type="tts"tts_config.textstringYes (for tts)Text to synthesizetts_config.audio_man_idstringYes (for tts)Voice IDtts_config.speednumberNoSpeech speed: 0.5-2, default 1tts_config.pitchnumberNoPitch, default 1audio_file_idstringYes (for audio)Audio file ID from File Management ( data.file_id) when audio_type="audio". Supports mp3, m4a, wavvolumeintNoVolume: 1-100, default 100

Response example:

json

{
  "trace_id": "8d10659438827bd4d59eaa2696f9d391",
  "code": 0,
  "msg": "success",
  "data": "9499ed79995c4bdb95f0d66ca84419fd"
}

Response field description:

FieldDescriptioncodeResponse status codemsgResponse messagedataVideo ID for subsequent polling

Query Task List#

Get a list of lip-syncing tasks.

http

POST /open/v1/video_lip_sync/list
access_token: {{access_token}}
Content-Type: application/json

Request body:

json

{
  "page": 1,
  "page_size": 10
}

Request field description:

ParameterTypeRequiredDescriptionpageintNoPage number, default 1page_sizeintNoPage size, default 10

Response example:

json

{
  "trace_id": "8d10659438827bd4d59eaa2696f9d391",
  "code": 0,
  "msg": "success",
  "data": {
    "list": [
      {
        "id": "9499ed79995c4bdb95f0d66ca84419fd",
        "status": 20,
        "progress": 100,
        "msg": "success",
        "video_url": "https://res.chanjing.cc/xxx/lip-sync/9499ed79995c4bdb95f0d66ca84419fd.mp4",
        "preview_url": "https://res.chanjing.cc/xxx/lip-sync/9499ed79995c4bdb95f0d66ca84419fd.jpg",
        "duration": 300,
        "create_time": 1738636800
      }
    ],
    "page_info": {
      "page": 1,
      "size": 10,
      "total_count": 1,
      "total_page": 1
    }
  }
}

Response field description:

First-level FieldSecond-level FieldDescriptioncodeResponse status codemsgResponse messagedataResponse datalistTask listid: Video IDstatus: Task status (0-pending, 10-generating, 20-success, 30-failed)progress: Progress 0-100msg: Task messagevideo_url: Video download URLpreview_url: Cover image URLduration: Video duration in mscreate_time: Creation time (unix timestamp)page_infoPagination info

Query Task Detail#

Poll the following API to check task status until completed.

http

GET /open/v1/video_lip_sync/detail
access_token: {{access_token}}

Query params:

ParameterDescriptionidVideo ID

Example: GET /open/v1/video_lip_sync/detail?id=9499ed79995c4bdb95f0d66ca84419fd

Response example:

json

{
  "trace_id": "8d10659438827bd4d59eaa2696f9d391",
  "code": 0,
  "msg": "success",
  "data": {
    "id": "9499ed79995c4bdb95f0d66ca84419fd",
    "status": 20,
    "progress": 100,
    "msg": "success",
    "video_url": "https://res.chanjing.cc/xxx/lip-sync/9499ed79995c4bdb95f0d66ca84419fd.mp4",
    "preview_url": "https://res.chanjing.cc/xxx/lip-sync/9499ed79995c4bdb95f0d66ca84419fd.jpg",
    "duration": 300,
    "create_time": 1738636800
  }
}

Response field description:

First-level FieldSecond-level FieldDescriptioncodeResponse status codemsgResponse messagedataResponse dataidVideo IDstatusTask status: 0-pending, 10-generating, 20-success, 30-failedprogressProgress 0-100msgTask messagevideo_urlVideo download URLpreview_urlCover image URLdurationVideo duration in mscreate_timeCreation time (unix timestamp)

Callback Notification#

When a callback URL is provided, the system will send a POST request when the task completes:

json

{
  "trace_id": "8d10659438827bd4d59eaa2696f9d391",
  "code": 0,
  "msg": "success",
  "data": {
    "id": "9499ed79995c4bdb95f0d66ca84419fd",
    "status": 20,
    "progress": 100,
    "msg": "success",
    "video_url": "https://res.chanjing.cc/xxx/lip-sync/9499ed79995c4bdb95f0d66ca84419fd.mp4",
    "preview_url": "https://res.chanjing.cc/xxx/lip-sync/9499ed79995c4bdb95f0d66ca84419fd.jpg",
    "duration": 300,
    "create_time": 1738636800
  }
}

Scripts#

本 Skill 提供脚本（ skills/chanjing-avatar/scripts/），与 chanjing-credentials-guard 使用同一配置文件（ ~/.chanjing/credentials.json）获取 Token。

脚本说明get_upload_url.py获取上传链接，输出 sign_url、 mime_type、 file_id``upload_file.py上传本地文件，轮询 file_detail 直到就绪后输出 file_id``create_task.py创建对口型任务（TTS 或音频驱动），输出视频任务 idpoll_task.py轮询任务直到完成，输出 video_url

示例（在项目根或 skill 目录下执行）：

bash

# 1. 上传驱动视频，得到 video_file_id
VIDEO_FILE_ID=$(python skills/chanjing-avatar/scripts/upload_file.py --service lip_sync_video --file ./my_video.mp4)

# 2. 创建 TTS 对口型任务（需先通过 list_common_audio 获取 audio_man_id）
TASK_ID=$(python skills/chanjing-avatar/scripts/create_task.py \
  --video-file-id "$VIDEO_FILE_ID" \
  --text "君不见黄河之水天上来" \
  --audio-man-id "C-f2429d07554749839849497589199916")

# 3. 轮询直到完成，得到视频下载链接
python skills/chanjing-avatar/scripts/poll_task.py --id "$TASK_ID"

音频驱动时：先上传音频得到 audio_file_id，再 create_task.py --video-file-id <id> --audio-file-id <audio_file_id>。

Response Status Code Description#

CodeDescription0Response successful10400AccessToken verification failed40000Parameter error40001Exceeds RPM/QPS limit50000System internal error