Skill Capabilities
AI-powered video analysis built on OpenClaw — teaching cameras to understand the world
Intelligently identify packages in surveillance footage with real-time detection of appearance, movement, and abandonment.
Core Capabilities
- Support for multiple package types (boxes, bags, envelopes, etc.)
- Real-time detection of package appearance, movement, and abandonment
- Automatic alerts for unattended packages
- Configurable alert rules and push notifications
- Use cases: delivery lockers, reception, warehouses
Person Detection
Deep learning-based person detection and recognition with real-time positioning and identification in multi-person scenarios.
Core Capabilities
- Multi-person simultaneous detection and tracking
- Body attribute analysis (clothing, posture, etc.)
- Cross-camera tracking of the same person
- Custom detection zones and alert rules
Video Search
Quickly search target clips in massive video archives using visual features, supporting image-based search and target backtracking.
Core Capabilities
- Image-based search to quickly locate target appearances
- Search by person, vehicle, object categories
- Timeline visualization of search results
- Multi-channel parallel video search
- Fast search through massive video archives
Text-based Video Search
Search video content using natural language descriptions — type text to find matching video clips.
Core Capabilities
- Natural language input with semantic-level video matching
- Support for scene and behavior description queries
- Cross-modal retrieval combining vision and language models
- Results ranked by relevance
Face Recognition
Face registration and real-time comparison for access control, attendance, and VIP identification scenarios.
Core Capabilities
- Face registration database with batch import
- Real-time face comparison and identification
- Liveness detection to prevent photo/video spoofing
- Support for single verification and multi-person search
Voice Video Search
Search video content via voice commands with speech-to-text conversion for semantic-level video matching.
Core Capabilities
- Automatic speech-to-text conversion
- Semantic understanding for video clip matching
- Chinese and English speech recognition
- Ideal for patrol and command center hands-free scenarios
Smoke Detection
Real-time smoke detection in surveillance footage for early fire warning.
Core Capabilities
- Real-time smoke detection with second-level response
- Distinguish smoke from fog and steam interference
- Support for indoor and outdoor scenarios
- Multi-channel alerts (email, webhook, SMS)
Vehicle Detection
Identify vehicles and detect theft, damage, or collision anomalies.
Core Capabilities
- Vehicle detection and license plate recognition
- Vehicle exterior damage detection (scratches, dents, collisions)
- Abnormal parking and suspicious vehicle alerts
- Vehicle entry/exit records and trajectory tracking
Abnormal Behavior Detection
Intelligently identify fighting, loitering, wandering and other abnormal behaviors with timely alerts.
Core Capabilities
- Automatic conflict behavior recognition
- Loitering and wandering detection
- Crowd density anomaly alerts
- Custom abnormal behavior rules
Try KamiClaw Skill Now
Sign up to get free credits and quickly integrate intelligent video security detection
Sign Up Free