Child Emotion Recognition Analyzer

Problem

In homes, kindergartens, and care centers, caregivers may not be able to watch surveillance footage continuously or respond quickly when a child shows crying, anger, fear, or distress. This skill turns surveillance images or video clips into a practical agent workflow for detecting negative emotions, issuing alerts, and keeping structured reports.

How It Works

• Input handling: accepts local files or network URLs, including jpg/jpeg/png/mp4/avi/mov, with a 10MB file limit.
• Recognition: runs the built-in script scripts.child_emotion_recognition_analysis to analyze facial expressions, vocal cues, and body movements in the supplied media.
• Output: returns a structured report with detected emotion signals, risk notes, recommendations, and a report link.
• History lookup: uses the --list parameter to query historical reports from the cloud API rather than local memory or manually maintained notes.

Boundaries

The results are reference information only and do not replace adult supervision. In urgent situations, on-site response is still required. The skill does not create temporary scripts and does not ask users for identity parameters; identity association is handled internally. It is suitable for tasks explicitly about child emotion recognition, crying detection, emotion alerts, or historical report review.