Exist

@Exist what correlations have you found between my sleep quality and productivity over the last month?

Based on your Exist data, here are three notable correlations from the past 30 days: (1) Strong positive correlation (0.72) between sleep duration >7 hours and next-day productive time — you averaged 4.2 more productive hours on well-rested days. (2) Moderate negative correlation (-0.54) between late bedtimes (after midnight) and morning mood ratings. (3) Weak positive correlation (0.31) between step count and sleep quality score. These patterns suggest sleep consistency may be your highest-leverage factor for productivity.

This example showcases Exist's core strength: surfacing data relationships you might miss manually. Requires the user to have been tracking both attributes (sleep + productivity metrics) for at least 2-3 weeks to generate meaningful correlations. The MCP returns correlation coefficients and date ranges; the AI interprets significance.

@Exist show me all the health and productivity attributes I'm currently tracking, grouped by category

You're tracking 18 attributes across 4 categories: **Health** (sleep hours, sleep quality, resting heart rate, steps, weight), **Productivity** (productive time, distractions, tasks completed), **Mood** (overall mood, energy level, stress), **Social** (time with friends, messages sent). Five attributes are auto-synced from integrations (Fitbit for sleep/steps, RescueTime for productive time), while 13 require manual daily input. Your longest tracking streak is 87 days for sleep hours.

Demonstrates the metadata retrieval capability — useful for auditing what data you're collecting before running analyses. The MCP returns attribute names, types, and sync sources; it does NOT return actual values (use a different tool for time-series data). Helps users understand their tracking coverage before requesting correlations.

@Exist draft a weekly review summarizing my top 3 habit patterns from the attribute templates you have available

Here's your week of Jan 6-12 by the numbers: **Sleep Consistency** — You hit your 7-hour target 5/7 nights, with bedtime variance under 45 minutes (your best week in a month). **Movement** — Daily steps averaged 8,200, up 18% from your December baseline; Friday's 12K spike correlates with your reported high energy. **Focus Blocks** — Productive time peaked Monday-Wednesday (avg 5.1 hrs/day) but dropped Thursday-Friday (2.8 hrs) as meeting load increased. One pattern to watch: your 3 lowest-sleep nights all preceded high-distraction days.

This synthesis example shows how the AI can combine attribute templates with correlation data to generate narrative insights. The MCP provides raw metrics and relationships; the AI does the storytelling. Requires OAuth2 scope for both reading user attributes and accessing correlation endpoints. Most valuable when the user tracks 5+ attributes regularly.

Exist attribute tracking for distributed freelancer capacity planning

An 8-person freelance collective shares client projects but struggles to distribute work fairly when someone's overloaded. Each member tracks daily work hours, stress levels, and project count in Exist. The collective's coordinator uses Switchy to pull attribute data weekly and spot who's trending toward burnout before they ghost a deadline. The Get User Attributes tool surfaces metadata without exposing raw daily entries, so members control granularity. This works if your collective already has a culture of transparency around workload—if people fake their numbers to seem available, the data is worthless. The threshold: below 5 people, a Slack check-in is faster; above 12, you need real workforce management software. Use Exist when you're in the middle zone where informal check-ins break down but formal tools feel like overkill.