menstrual-cycle-tracking_skill

Overview

This skill records and analyzes menstrual cycle entries sent in natural language starting with the trigger phrase "record menstrual cycle." It extracts dates, cycle day, flow, pain level, mood, and symptoms, saves structured JSON records, and produces simple trend insights from historical data. The focus is reliable extraction and lightweight health trend summarization.

How this skill works

When a message begins with "record menstrual cycle," the skill parses the raw text to identify date, cycle_day (or infers it from the stored cycle start), flow (e.g., light/medium/heavy), pain (0–10), mood, and symptoms. It writes an entry to a JSON file named menstrual_cycle_tracking_records.json with timestamp, raw_text, and a structured object. Periodic analysis routines summarize trends such as average cycle length, typical pain levels, and changes in flow over time.

When to use it

• User wants to log a daily or occasional menstrual update in natural language.
• When extracting structured cycle data from free-text notes for tracking.
• To build simple historical summaries like average cycle length or pain trends.
• When users don’t provide explicit cycle day and you need to infer it from previous entries.
• To persist entries locally as JSON for backup or later analysis.

Best practices

• Require the message to start with the trigger phrase to avoid accidental parsing.
• Normalize flow to a small set (light, medium, heavy) and pain to 0–10 for consistency.
• Infer cycle_day only when a clear cycle start is present in history; otherwise ask a clarifying question.
• Store both raw_text and structured fields to enable iterative improvement of parsers.
• Run weekly or monthly analytics jobs to surface trends and flag anomalies for review.

Example use cases

• User: "record menstrual cycle: second day, very heavy flow, severe cramps, mood anxious." Skill extracts cycle_day=2, flow=heavy, pain≈8, mood=anxious and saves a JSON record.
• Automatic inference of cycle_day when user logs a start date and later logs days without specifying day number.
• Generate a monthly report showing average pain and frequency of heavy flow days.
• Detect a sudden increase in pain or irregular cycle length and prompt the user to consult a clinician.
• Backup all entries to the JSON file for migration or archival purposes.