weryai-video-toolkits（weryai 视频工具包）

WeryAI Video Toolkits

WeryAI video tools, video editing, video post-processing, subtitle translate, subtitle erase, watermark remove, face change, lip sync, background remove, video extend, anime replace, magic style transfer, video upscale.

WeryAI video tools for video editing and video post-processing on existing videos.

Use this skill for official WeryAI video editing, video post-processing, and WeryAI video tools on existing video/media sources. Use when you need WeryAI video tools for subtitle translate, subtitle erase, watermark remove, face change, lip sync, background remove, video extend, anime replace, magic style transfer, or video upscale. This skill is intentionally strict about secret declaration and input safety: the only runtime secret is WERYAI_API_KEY, and media inputs support http/https URLs plus local/file sources that are uploaded first.

This is not a text-to-video generator. Use it when the user wants to edit, transform, clean up, translate, upscale, or extend an existing video with WeryAI rather than create a brand-new video from a prompt.

Dependencies: scripts/video_toolkits.js at the skill package root (alongside SKILL.md), WERYAI_API_KEY, and Node.js 18+. No other skills are required.

Example Prompts

• Remove the watermark from this HTTPS video with WeryAI.
• Translate the subtitles in this video to English and give me the final video URL.
• Use WeryAI to replace the face in this video with this reference photo.
• Extend this clip by 5 seconds in anime style.
• Lip-sync this talking-head video to the provided audio track.

Quick Summary

• Main jobs: anime-replace, background-remove, extend, face-change, lips-change, magic-style, subtitle-erase, subtitle-translate, upscaler, watermark-remove
• Main inputs: video_url plus tool-specific parameters such as image_url, audio_url, target_language, rect_vo_list, style, resolution, duration
• Main trust signals: dry-run support, explicit per-tool validation, parameter guidance by intent, paid-run warning, WERYAI_API_KEY alignment with other WeryAI skills

Authentication and first-time setup

Before the first real processing run:

1. Create a WeryAI account.
2. Open the API key page at https://www.weryai.com/api/keys.
3. Create a new API key and copy the secret value.
4. Add it to the required environment variable WERYAI_API_KEY.
5. Make sure the WeryAI account has available balance or credits before paid processing.

OpenClaw-friendly setup

• This skill already declares WERYAI_API_KEY in metadata.openclaw.requires.env and primaryEnv.
• After installation, if the runtime asks for required environment variables, paste the key into WERYAI_API_KEY.
• If you are configuring the runtime manually, export it before running commands:

export WERYAI_API_KEY="your_api_key_here"

Quick verification

Use one safe check before the first paid run:

node scripts/video_toolkits.js tools
node scripts/video_toolkits.js wait --tool background-remove --json '{"video_url":"https://example.com/video.mp4"}' --dry-run

• tools confirms the local CLI is available and shows the supported tool registry.
• --dry-run validates the request shape without calling WeryAI or spending credits.

Prerequisites

• WERYAI_API_KEY must be set before running video_toolkits.js for paid calls.
• Node.js >=18 is required because the runtime uses built-in fetch.
• video_url, image_url, and audio_url can be http/https URLs or local/file sources. Local/non-http(s) sources are uploaded first.
• Real submit, wait, and status commands can consume WeryAI credits or depend on existing paid tasks.

Security and API hosts

• WERYAI_API_KEY: Treat it as a secret. Configure it only in the runtime environment; never write the secret value into the skill files.
• This skill supports directly passing local file paths. If a local file path is provided, the runtime will automatically upload the local file to the WeryAI server for processing.
• Optional override WERYAI_BASE_URL defaults to https://api.weryai.com. Only override it with a trusted host.
• For higher assurance, run paid jobs in a short-lived shell or isolated environment, and review scripts/video_toolkits.js before production use.

Tool Routing

Route user intent to the narrowest tool:

• Replace or move an object in anime style -> anime-replace
• Remove background or swap it to a flat color -> background-remove
• Continue a short clip -> extend
• Swap a face in a source video -> face-change
• Lip-sync video to audio -> lips-change
• Apply a preset visual style -> magic-style
• Remove subtitles or burnt-in text -> subtitle-erase
• Translate subtitles -> subtitle-translate
• Improve resolution -> upscaler
• Remove watermark -> watermark-remove

If the user asks for text-to-video or image-to-video generation from scratch, use weryai-video-generator instead of this skill.

Parameter Guidance

Guide the user progressively. Ask only for the smallest missing set of parameters required for the selected tool.

Recommended guidance pattern

Use short operator-style guidance like this:

• General help: When the user asks "how to use this skill", DO NOT paste raw shell commands. Instead, explain the capabilities in natural language and give 2-3 prompt examples.

Defaults you may apply safely

• anime-replace: type=replace, resolution=720p
• background-remove: background_color=BLACK
• extend: style=anime, resolution=720p, duration=5
• magic-style: video_style=anime_style_3
• upscaler: resolution=1080p
• subtitle-erase and watermark-remove: omit rect_vo_list to let WeryAI auto-detect regions

Parameters that usually require user confirmation

• subtitle-translate.target_language
• face-change.image_url
• lips-change.audio_url
• precise rect_vo_list region coordinates when auto-detection is not good enough

When to ask follow-up questions

• Ask for target_language when the user says “translate subtitles” but does not name the destination language.
• Ask for the reference image_url when the user wants face replacement but has not supplied a face image.
• Ask for audio_url when the user wants lip-sync but has not supplied audio.
• Ask for rect_vo_list only when the user wants precise control over subtitle or watermark regions.
• Ask about style, duration, or resolution for extend only if the user wants something different from the default continuation setup.

Tool Matrix

Read references/video-tools-matrix.md when you need the exact required fields, enum values, or defaults for a tool.

Preferred Commands

# Remove background with default BLACK fill
node scripts/video_toolkits.js wait --tool background-remove --json '{"video_url":"https://example.com/video.mp4"}'

# Replace or move an object in anime style
node scripts/video_toolkits.js wait --tool anime-replace --json '{"video_url":"https://example.com/video.mp4","image_url":"https://example.com/ref.jpg","type":"replace","resolution":"720p"}'

# Poll an existing task
node scripts/video_toolkits.js status --task-id <task-id>

Workflow

1. Identify whether the user wants video editing/post-processing or brand-new video generation.
2. Route the request to one tool ID.
3. Collect video_url and only the extra parameters required by that tool.
4. Apply supported defaults where safe.
5. Show the final tool, parameters, and URLs before the paid run if the request is ambiguous or expensive.
6. Use --dry-run when you need to verify the payload locally first.
7. Use wait by default to deliver final processed video URLs in the same turn.
8. Enforce bounded polling with a maximum timeout of 30 minutes (1800 seconds); do not run unbounded loops.
9. If timeout is reached, return the taskId to the user and ask if they want you to check the status again. Do NOT show the raw node status command to the user; use it internally to check the status when asked.
10. Use submit only when the user explicitly wants task creation without polling.
11. Use status when the user already has a task_id.

Input Rules

• video_url, image_url, and audio_url can be http/https URLs or local/file sources.
• Do not invent undocumented fields.
• Use normalized rect_vo_list coordinates between 0 and 1.
• For subtitle or watermark removal, omit rect_vo_list when auto-detection is acceptable.
• For subtitle translation, use standard language codes such as en, zh, ja, ko, fr, de, es.

Output

All commands print JSON to stdout. Successful results can include:

• taskId, taskIds, batchId
• taskStatus
• videos
• requestSummary
• errorCode, errorMessage

User-facing delivery requirement:

• If video URLs are available, return at least one playable Markdown link (for example [Video](https://...)). If multiple videos are generated, render all of them using markdown links consecutively.
• Alongside video output, include key parameters when available: tool, style, duration, resolution, type, target_language.
• Do not use taskId as the sole deliverable unless the user explicitly requested task creation without waiting.
• If timeout is reached before completion, return the taskId to the user and ask if they want you to check the status again. Do NOT show the raw node status command to the user; use it internally to check the status when asked.

The task is done when:

• local validation passes without CLI-side errors,
• submit returns a valid task ID or batch ID,
• or wait reaches a terminal result with at least one processed video URL,
• or status returns a clear in-progress or terminal state,
• and the output makes it explicit whether final video URLs are present.

Constraints

• Do not treat this skill as a text-to-video generator.
• Do not ask the user to choose raw API fields when a safe default already fits the request.
• Do not use unsupported file types for video_url, image_url, or audio_url; local sources are uploaded before the toolkit request.
• Do not re-run paid processing casually because each submit or wait call can create a new paid task.
• Do not broaden this skill beyond the documented WeryAI video-tools API surface.

Re-run behavior

• submit and wait are not idempotent.
• status and tools are safe to re-run.

References

• Video tools parameter matrix: references/video-tools-matrix.md
• Official documentation index: WeryAI llms.txt
• Anime replace API: Submit Video Anime Replace Task