← 返回 Skills 市场
WeryAI Video Toolkits
作者
WeryAI-Developer
· GitHub ↗
· v0.1.3
· MIT-0
172
总下载
0
收藏
0
当前安装
4
版本数
在 OpenClaw 中安装
/install weryai-video-toolkits
功能描述
Process and edit existing videos using WeryAI video toolkits. Use when the user needs to remove video backgrounds, replace backgrounds, apply anime style tra...
安全使用建议
This package largely does what it says: a Node CLI that uploads videos/images/audio and calls the WeryAI API to edit videos. Before installing or running: 1) Do not paste your WERYAI_API_KEY into files — set it in the runtime environment as instructed. 2) Confirm whether your runtime will prompt for WERYAI_BASE_URL; do not set this to an untrusted or unknown host (the code will send uploads and API calls to that URL). 3) Be aware that supplying local file paths causes the CLI to read and upload those files to the API — avoid pointing it at sensitive local files. 4) The CLI forwards any webhook_url you provide to the API; avoid passing webhook URLs you don't control or trust to prevent unexpected data exfiltration. 5) Use the provided dry-run mode to verify request shapes before spending credits or uploading real files, and consider running the tool in an isolated/short-lived environment when doing paid/risky runs. 6) If you need higher assurance, review the bundled scripts (especially the client/upload code) or run the tool behind network controls before providing secrets or local files.
功能分析
Type: OpenClaw Skill
Name: weryai-video-toolkits
Version: 0.1.3
The skill bundle is a legitimate and well-structured toolkit for interacting with the WeryAI Video API. It provides comprehensive wrappers for video editing tasks such as upscaling, background removal, and style transfer. While the skill includes high-risk capabilities like local file uploading and network access, these are essential for its stated purpose and are protected by safeguards including file extension whitelisting (e.g., in `scripts/vendor/weryai-core/upload.js`), size limits, and security warnings when connecting to non-official domains. No evidence of intentional data exfiltration, malicious execution, or obfuscation was found.
能力评估
Purpose & Capability
Name, description, and included scripts align with a WeryAI video-editing client: node CLI, many tool-specific endpoints, input validation, dry-run support, and local-file upload. However the declared required environment variables are inconsistent with the prose: SKILL.md describes WERYAI_BASE_URL as an optional override (defaults to https://api.weryai.com), but the registry metadata and top-level requirements list WERYAI_BASE_URL as required. Requiring a base URL is disproportionate for a client that normally uses the vendor's default endpoint.
Instruction Scope
SKILL.md instructs the agent to run the included node CLI and to supply WERYAI_API_KEY; it supports dry-run and uploading local files. The runtime instructions and code limit scope to WeryAI API interactions and local-file validation/upload. Notably, the code accepts a webhook_url parameter (passed through to the API) and automatically uploads local file paths to the remote API — both are legitimate for this tool but expand the data flow surface beyond simple API calls (local files will be transmitted to the remote service; webhooks could direct results to third-party endpoints).
Install Mechanism
No installer or remote downloads are used; this is an instruction-only package with all runtime code included. It requires Node.js to execute the bundled scripts. No fetch-from-arbitrary-URL installation steps were found.
Credentials
Primary credential WERYAI_API_KEY is appropriate and expected. However, the registry lists WERYAI_BASE_URL as a required environment variable while SKILL.md documents it as optional and defaults to the official API host. Requiring a base URL is unnecessary and could cause confusion or misconfiguration. Because the code uses the base URL for uploads and API calls, setting it to an untrusted host would redirect uploads and API traffic (including any local files) away from the official service. Also the runtime accepts a user-provided webhook_url that will be forwarded to the API — this can cause results or callbacks to be delivered to arbitrary endpoints if the user supplies them.
Persistence & Privilege
The skill does not request always:true, does not claim to persist beyond its own files, and does not modify other skills or global agent settings. Autonomous invocation remains allowed (default) but that is expected for skills. There are no special privilege escalations declared.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install weryai-video-toolkits - 安装完成后，直接呼叫该 Skill 的名称或使用
/weryai-video-toolkits触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v0.1.3
- Adds internal vendor code for WeryAI core and video functionality in the scripts/vendor directory.
- Introduces support for processing local file paths; local files are now auto-uploaded to WeryAI.
- Expands required environment variables to allow API host override with WERYAI_BASE_URL.
- Improves documentation regarding local file handling and WERYAI_BASE_URL usage.
- No changes to main prompts, tool selection, or default behaviors for video toolkits.
v0.1.2
- Major update: Skill refocused from "video generation" to "video editing & post-processing" with toolkits.
- Rebranded and simplified; removed video generation scripts and references, added concise toolkit-focused logic.
- Now supports operations like background removal, watermark removal, face change, lip-sync, subtitle translation/erasure, anime style transfer, upscaling, and extension of existing videos.
- Only requires two files: core toolkit script and a video tools matrix reference.
- Tightened security and parameter guidance; only `WERYAI_API_KEY` as a secret.
- Updated user guidance for tool selection, progressive parameter collection, and safe usage examples.
v0.1.1
**Major update: Skill refactored and renamed to shift from video editing to video generation with WeryAI.**
- Renamed the skill from video toolkits/post-processing to "weryai-video-generator", now focused on generating videos from text, images, or storyboards.
- Added extensive scripts for new CLI workflows: model capability lookup, payload previews (dry-run), and multi-modal input support.
- Default model is now Seedance 2.0 with sensible parameters (5s, 720p, 9:16, audio enabled), but model switching and parameter overrides are supported.
- Added model/parameter guidance patterns and mappings for common use cases (e.g., aspect ratio by platform, prompt to parameter translation).
- Updated documentation to reflect new usage patterns, CLI commands, parameter safety, and user guidance.
- Removed legacy files and code for post-processing/editing workflows; exclusively supports video generation and status checking.
v0.1.0
Initial release of WeryAI Video Toolkits skill.
- Provides video post-processing tools: subtitle translate/erase, watermark remove, face change, lip sync, background remove, video extend, anime replace, magic style transfer, and video upscaler.
- Requires WERYAI_API_KEY and Node.js 18+; works with public HTTPS media URLs only.
- Strict input validation and paid-run awareness; supports dry-run for request validation.
- Includes clear setup, authentication, and usage instructions.
- Guides users to supply minimum required parameters per tool; offers default values and follow-up prompting for missing info.
- Recommends users switch to a different skill for text-to-video generation.
元数据
常见问题
WeryAI Video Toolkits 是什么？
Process and edit existing videos using WeryAI video toolkits. Use when the user needs to remove video backgrounds, replace backgrounds, apply anime style tra... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 172 次。
如何安装 WeryAI Video Toolkits？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install weryai-video-toolkits」即可一键安装，无需额外配置。
WeryAI Video Toolkits 是免费的吗？
是的，WeryAI Video Toolkits 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
WeryAI Video Toolkits 支持哪些平台？
WeryAI Video Toolkits 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 WeryAI Video Toolkits？
由 WeryAI-Developer（@weryai-developer）开发并维护，当前版本 v0.1.3。
推荐 Skills