← Back to Skills Marketplace
85
Downloads
0
Stars
0
Active Installs
7
Versions
Install in OpenClaw
/install next-video-gen
Description
豆包 Next Video Gen AI 视频/图片生成 via 火山引擎方舟平台。支持文生图、文生视频、图生视频、素材生视频，集成 Seedance 系列模型。
Usage Guidance
This package appears to be what it claims (a Volcengine Seedance-based media generator) and only needs your DOUBAO_API_KEY, but there are a few packaging inconsistencies you should check before installing:
- Verify API key scope and rotate/revoke it if you later remove the skill. Only provide DOUBAO_API_KEY if you trust the source.
- The registry metadata omitted the required env/binary declarations that appear in SKILL.md and scripts (node required, DOUBAO_API_KEY required). Treat this as a packaging oversight but confirm you are comfortable giving the skill a key.
- Installer (bin/cli.js) runs local commands to find the skills directory and will copy files into ~/.openclaw/skills (or a directory you specify). Review the files locally (SKILL.md, scripts/, bin/) before running the installer and avoid running unreviewed installers as root.
- The scripts will fetch user-supplied URLs (images/videos/audio) and save generated outputs to ~/Videos/next-video-gen/ by default — be mindful of what URLs you provide and where output is written.
- The installer uses child_process (execSync/spawnSync) for detection; this is normal for installers but means it will execute shell commands locally — inspect the script if you have doubts.
If you want higher assurance, ask the publisher to fix the metadata mismatch (declare node and DOUBAO_API_KEY in the registry fields) or run the scripts in an isolated environment (container or dedicated VM) and inspect the code prior to providing your API key.
Capability Analysis
Type: OpenClaw Skill
Name: next-video-gen
Version: 1.0.6
The skill bundle is a legitimate tool for generating AI images and videos using the Volcengine Ark (Seedance) API. The scripts (seedance-gen.sh, run-gen.js) and utilities (http-request.js, download-file.js) are well-structured, use standard Node.js modules for networking, and interact only with official API endpoints. The installer (bin/cli.js) follows standard practices by detecting skill directories and optionally configuring the required API key in shell profile files with user consent.
Capability Tags
Capability Assessment
Purpose & Capability
Functionality (text→image/video, image→video, video→video) matches the description and all network calls target Volcengine Ark endpoints (ark.cn-beijing.volces.com). Requesting DOUBAO_API_KEY is appropriate for this purpose. However, the registry header lists no required env/bins while SKILL.md and the scripts require node and DOUBAO_API_KEY — this metadata mismatch is inconsistent.
Instruction Scope
SKILL.md and the scripts guide the agent to copy files into a skills directory, check dependencies, prompt for API key, call the Volcengine APIs, and download generated media to ~/Videos/next-video-gen/. The runtime instructions do not ask for unrelated system secrets or to read arbitrary local files, but the installer will probe for an OpenClaw skills path (running 'which/where' and optionally calling openclaw CLI) and will copy files into user directories. The scripts also accept user-supplied remote URLs (images/videos/audio) which will be fetched — expected but worth noting.
Install Mechanism
There is no formal install spec in registry metadata; instead an included CLI installer (bin/cli.js) copies files into the user's skills directory. All code is included in the package (no downloads from arbitrary third-party URLs during install). The installer uses child_process execSync/spawnSync to probe the environment (e.g., checking openclaw); this is expected for an installer but increases the surface you should trust. No remote install/extraction from unknown hosts was observed.
Credentials
Runtime requires only DOUBAO_API_KEY (primaryEnv) which is proportionate to calling Volcengine Ark APIs. However, the registry metadata at the top of the package listing states 'Required env vars: none' while SKILL.md metadata and code require DOUBAO_API_KEY — this discrepancy is misleading. The scripts also recommend/expect curl and jq (installer checks for them), but the Node scripts implement HTTP/downloads themselves, so the curl/jq requirement appears inconsistent.
Persistence & Privilege
Skill is not forced 'always' and does not request elevated or cross-skill configuration changes. The installer will copy files into a per-user skills directory and set execute permissions on its own scripts — normal for a skill installer.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install next-video-gen - After installation, invoke the skill by name or use
/next-video-gen - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.6
- 补充“字幕处理”说明：目前 API 无原生字幕参数，如需字幕可在 prompt 描述中注明字幕内容，由模型根据描述实现。
- 其余文档与功能保持一致，无破坏性变更。
v1.0.5
- 新增参数 --enable-audio，用于启用视频音频输出，替代原用法
- 文档中参数说明由 --audio 改为 --enable-audio，明确区分与参考音频（--audio <url>）
- 文档及脚本用法保持对旧版 --audio 参数的兼容说明
- 其余功能与接口无变化，仅参数命名与使用方式更规范
v1.0.4
- 改进文生视频音频参数，--audio 替换原 --no-audio；用法和参数文档均已更新
- 修正 run-gen.js 脚本，--audio 用法更直观，兼容带参和不带参形式
- 文档示例全部更新为新版 --audio 用法，便于理解
- 细化参数说明，清理旧用法描述（如 --no-audio）
- 保持 Bash 脚本兼容，未影响现有功能
v1.0.3
- Updated output examples and delivery standards to clarify that online links are only valid for a short time and users should download promptly.
- Changed wording in protocol and delivery sections to emphasize the temporary validity of generated image and video URLs.
- Added a prominent note explaining that online links contain temporary signatures, usually valid for 24 hours, and must be regenerated after expiry.
- No core parameter or usage changes; documentation improved for user clarity around file downloads.
v1.0.2
- Added Node.js script usage instructions for cross-platform support (run-gen.js).
- Clarified video/audio option (`--no-audio` now default "no", use `--no-audio false` to enable audio).
- Bash script is now documented as "legacy, compatible".
- Documented a new error code for Windows HTTPS issues and suggested retrying.
- Minor improvements to examples and default spec presentation in output.
v1.0.1
- 新增脚本 scripts/run-gen.js，为功能扩展做准备
- 其他未变更
v1.0.0
Initial release of next-video-gen:
- Provides AI-based image and video generation via Huoshan Ark Platform using Seedance models.
- Supports text-to-image, text-to-video, image-to-video, and video-to-video modes.
- Exposes configurable parameters for quality, aspect ratio, watermark, duration, and model selection.
- Includes a bash+node script for command-line generation with clear output protocol.
- Handles and reports progress, errors, and result details—including online links and local file locations.
- Guides users to configure required DOUBAO_API_KEY on first use.
Metadata
Frequently Asked Questions
What is Next Video Gen?
豆包 Next Video Gen AI 视频/图片生成 via 火山引擎方舟平台。支持文生图、文生视频、图生视频、素材生视频，集成 Seedance 系列模型。 It is an AI Agent Skill for Claude Code / OpenClaw, with 85 downloads so far.
How do I install Next Video Gen?
Run "/install next-video-gen" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Next Video Gen free?
Yes, Next Video Gen is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Next Video Gen support?
Next Video Gen is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Next Video Gen?
It is built and maintained by Venn (@vennduan); the current version is v1.0.6.
More Skills