freddy.

freddy is an MCP server that plugs your wearables straight into Claude, ChatGPT, and any AI agent. Think of it as giving your body a voice. You connect Oura, WHOOP, Garmin, Dexcom, Wahoo, Hevy—19 sources total—then just ask your AI the question you've been staring at dashboards trying to answer. "Why is my HRV tanking this week?" It pulls your real numbers and tells you training load climbed 34% while sleep stayed flat, baseline HRV drifted from 56 to 49ms over 30 days. That's not a dashboard. That's a conversation.

Here's the problem freddy solves: you have more biometric signal than any athlete in history—HRV, glucose, power, sleep stages, lifts, temperature, recovery scores—and every app gives you a number but none of them tell you why. Oura shows 78 readiness. Dexcom shows 147 mg/dL. WHOOP shows whatever. You're swimming in data and dying of thirst for answers. freddy connects once, gives you one MCP endpoint, and now your AI has context. It tracks 100+ metrics and the setup takes 90 seconds. Free to start, no card required.

This is the MCP play everyone in AI is going to be doing—connecting real-time personal data to agents instead of making them guess. Your body is generating signal 24/7 and every other app asks you what you want for breakfast. freddy asks what you actually need to know.