← Back to Skills Marketplace
324
Downloads
0
Stars
1
Active Installs
2
Versions
Install in OpenClaw
/install video-motion-detector
Description
视频动作检测与分析。用于检测视频中的特定动作序列，如武术动作、舞蹈动作、人体姿态变化等。当用户要求检测视频中的动作、分析动作是否完成、验证特定动作序列时触发此技能。支持输入视频URL或本地视频文件，检测完成后通过飞书通知结果。
Usage Guidance
Before installing: ask the publisher to clarify how 'read' and 'message' tools are provided and where model inference runs (locally or remote). Confirm which binaries are required (yt-dlp/you-get/ffmpeg) and whether the platform will supply them. Ask how Feishu credentials or webhook URLs are configured and stored — the SKILL.md shows a Feishu target but the skill metadata doesn't request any token/webhook. Be cautious about downloading videos that require login or contain sensitive content; if you must run this, avoid supplying broad system credentials and prefer sandboxed execution. If the publisher cannot explain the missing env/binary declarations and the notification/auth flow, treat the skill as risky and avoid installing it.
Capability Analysis
Type: OpenClaw Skill
Name: video-motion-detector
Version: 1.0.1
The skill bundle provides instructions for an AI agent to perform video motion analysis by downloading videos (using yt-dlp/you-get), extracting frames (using ffmpeg), and analyzing them with vision models. The workflow concludes by sending results via a messaging tool (Feishu). All described actions and commands in SKILL.md are directly aligned with the stated purpose, and there is no evidence of malicious intent, unauthorized data access, or obfuscation.
Capability Assessment
Purpose & Capability
The SKILL.md describes downloading videos (you-get, yt-dlp), frame extraction (ffmpeg), visual model inference, and sending Feishu notifications. Those capabilities align with a video-motion-detector, but the registry metadata declares no required binaries or credentials—so the skill's declared requirements do not match what the instructions actually need.
Instruction Scope
Instructions explicitly tell the agent to download arbitrary video URLs, extract frames, run visual-model analysis, and send results to Feishu using a 'message' tool. The doc references 'read' and 'message' tools and platform login needs but does not define them or specify how credentials/cookies are obtained or stored. That leaves ambiguity about where video data and frames are processed and where results are sent.
Install Mechanism
This is an instruction-only skill with no install spec or code files, which minimizes direct disk or executable installation risk. However, it does assume existence of third‑party binaries (yt-dlp/you-get/ffmpeg) that the metadata does not declare.
Credentials
No environment variables or credentials are declared, yet the skill workflow requires sending Feishu messages (which normally needs a webhook/token/user ID) and may require login cookies or credentials for some video platforms. The absence of declared secrets/credentials is an unexplained mismatch.
Persistence & Privilege
The skill is not always-on and has no install-time persistence or claims to modify other skills or system-wide settings. Autonomous invocation is allowed by default but not combined with elevated permanence here.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install video-motion-detector - After installation, invoke the skill by name or use
/video-motion-detector - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
Initial release of video-motion-detector.
- Detects specific motion sequences in videos, such as martial arts moves, dance, or posture changes.
- Supports input via video URL or local video files.
- Provides detailed steps for video download, frame extraction, and motion analysis.
- Allows detection of motion order, simultaneity, and repetition per user request.
- Sends structured detection results via Feishu notifications.
v1.0.0
Initial release of video-motion-detector.
- Detects and analyzes specific motion sequences in videos (e.g., martial arts, dance moves, posture changes).
- Supports video input via URL or local file, with platform-specific download instructions.
- Provides detailed frame extraction and analysis guidelines to ensure accurate detection of rapid or complex movements.
- Supports ordered, simultaneous, and count-based action sequence verification.
- Sends results and structured detection summaries to Feishu (Lark) users upon completion.
Metadata
Frequently Asked Questions
What is Na He?
视频动作检测与分析。用于检测视频中的特定动作序列，如武术动作、舞蹈动作、人体姿态变化等。当用户要求检测视频中的动作、分析动作是否完成、验证特定动作序列时触发此技能。支持输入视频URL或本地视频文件，检测完成后通过飞书通知结果。 It is an AI Agent Skill for Claude Code / OpenClaw, with 324 downloads so far.
How do I install Na He?
Run "/install video-motion-detector" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Na He free?
Yes, Na He is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Na He support?
Na He is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Na He?
It is built and maintained by hena1985 (@hena1985); the current version is v1.0.1.
More Skills