← Back to Skills Marketplace
minimax-video
by
silingyuan0
· GitHub ↗
· v1.0.0
· MIT-0
183
Downloads
1
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install minimax-video-generate
Description
MiniMax 视频生成技能 - 支持文生视频(Text-to-Video)、图生视频(Image-to-Video)、首尾帧视频(Frame-to-Frame)、主体参考视频(Subject-to-Video)。使用模型 MiniMax-Hailuo-2.3，默认生成 5 秒 720P MP4 视频，自动下载保...
Usage Guidance
This skill appears to implement a reasonable MiniMax video client, but before installing: 1) note that the script requires MINIMAX_API_KEY and optional MINIMAX_REGION even though the registry metadata does not declare them — treat the metadata as incomplete. 2) Verify the API endpoints (api.minimaxi.com / api.minimax.io) are the legitimate service you expect. 3) Provide a scoped or disposable API key (monitor usage) rather than a long-lived high-privilege key. 4) Because the skill will make network requests and write downloaded video files, run it in an isolated environment if you have concerns and inspect network logs. 5) If you need stronger assurance, ask the publisher for provenance (homepage/repo) or request that required env vars be added to the registry metadata and for a checksum/signature for the script.
Capability Analysis
Type: OpenClaw Skill
Name: minimax-video-generate
Version: 1.0.0
The skill bundle provides a functional interface for the MiniMax Video Generation API, supporting text-to-video, image-to-video, and frame-based generation. The core logic in `scripts/video.py` correctly implements API authentication via environment variables and uses official MiniMax endpoints (api.minimaxi.com/api.minimax.io) for task submission and file downloads. No evidence of malicious intent, data exfiltration, or prompt injection was found.
Capability Assessment
Purpose & Capability
The name/description (MiniMax video generation) aligns with the included code and SKILL.md: the script posts to a MiniMax video-generation API and downloads resulting MP4s. Requiring an API key and region is reasonable for this purpose. However, the registry metadata declares no required environment variables or primary credential while SKILL.md and the script both require MINIMAX_API_KEY and optionally MINIMAX_REGION — an incoherence between manifest and implementation.
Instruction Scope
SKILL.md instructions and examples match the script: they instruct creating tasks, polling status, and downloading video files. The instructions explicitly use MINIMAX_API_KEY and MINIMAX_REGION and direct the agent to call external MiniMax endpoints and save files locally. There is no scope creep (no hidden file/system scanning or unrelated credential access) inside the provided instructions.
Install Mechanism
No install spec is provided (instruction-only with an included Python script). No external archives or third-party package downloads are executed as part of an installer, so the installation risk is low. The script uses the requests library but does not attempt to install packages itself.
Credentials
The script and SKILL.md legitimately require MINIMAX_API_KEY (and optionally MINIMAX_REGION). Those credentials are proportional to the stated functionality. The concern is the registry metadata lists no required env vars/primary credential — this mismatch could cause surprising runtime failures or indicate incomplete metadata. The skill only reads MINIMAX_API_KEY and MINIMAX_REGION; it doesn't access other secrets or system config paths.
Persistence & Privilege
The skill does not request persistent/always-on presence (always: false) and does not modify other skills or system-wide settings. It performs network calls and writes downloaded video files to disk, which is expected for its function but should be noted.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install minimax-video-generate - After installation, invoke the skill by name or use
/minimax-video-generate - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of the MiniMax 视频生成技能 (minimax-video):
- Supports text-to-video, image-to-video, frame-to-frame, and subject-to-video generation using the MiniMax-Hailuo-2.3 model.
- Default output is a 5-second 720P MP4 video with automatic local download.
- Provides functions for job submission, task status querying, polling/waiting, and file downloading.
- Includes multiple usage examples for all supported modes.
- Configuration options allow region selection and API key setup.
Metadata
Frequently Asked Questions
What is minimax-video?
MiniMax 视频生成技能 - 支持文生视频(Text-to-Video)、图生视频(Image-to-Video)、首尾帧视频(Frame-to-Frame)、主体参考视频(Subject-to-Video)。使用模型 MiniMax-Hailuo-2.3，默认生成 5 秒 720P MP4 视频，自动下载保... It is an AI Agent Skill for Claude Code / OpenClaw, with 183 downloads so far.
How do I install minimax-video?
Run "/install minimax-video-generate" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is minimax-video free?
Yes, minimax-video is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does minimax-video support?
minimax-video is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created minimax-video?
It is built and maintained by silingyuan0 (@silingyuan0); the current version is v1.0.0.
More Skills