Smart Infant Cry Analysis

What It Solves

When a baby cries, caregivers often have to infer whether the cause is hunger, tiredness, pain, discomfort, or distress, especially at night. This skill takes an uploaded baby-cry audio or video file, analyzes the likely cause, and presents the result as a readable report.

How It Works

• Input: Supports mp3, wav, mp4, avi, and mov files up to 10MB; remote URLs are downloaded by the API service, while local files are uploaded by the script.
• Analysis: Runs scripts/infant_cry_analysis.py to classify the cry into types such as hunger, tiredness, pain, physical discomfort, distress, and wet diaper.
• Output: Returns basic audio information, detected cause, confidence, care suggestions, and historical report links. Past reports are fetched from the cloud API and rendered as a Markdown table with report name, result, confidence, timestamp, and view link.

Boundaries and Caveats

The audio should be clear, complete, and preferably 5-30 seconds long, with low background noise. Results are intended for parenting reference and do not replace medical advice; if the baby cries continuously, shows fever, or appears uncomfortable, seek medical care promptly. The workflow should not aggregate local history files or skip the required open-id step before calling the API.