Plug your wearables into Claude, OpenClaw, and any AI
freddy is a personal health MCP server that bridges wearables, CGMs, power meters, and gym apps with AI agents like Claude, ChatGPT, and OpenClaw. Designed for athletes, biohackers, and anyone who tracks multiple health metrics, it transforms fragmented data into a single conversational interface that replaces scattered dashboards. Instead of toggling between Oura, WHOOP, and Garmin apps to piece together insights, users simply ask their AI questions like 'Why is my HRV tanking?' and get answers grounded in real, synced data from devices like Dexcom and Wahoo. freddy's core value is eliminating the gap between data collection and actionable insight, making health analytics as natural as talking to a coach, but with the depth of cross-source correlation that no single device app provides.
The modern athlete collects more signal than ever: sleep, HRV, glucose, power, lifts, and more from dedicated wearables and apps. Yet the apps that ship with these devices ask trivial questions like what you want for breakfast, while users are left staring at scattered scores and scores without understanding why. Three apps, twelve numbers, zero answers sums up the frustration: users open Oura to see a readiness score of 78, Dexcom showing 147 mg/dL with spikes, and Wahoo reporting FTP at 214 with a form of −18, but none tell you why your recovery is suffering. freddy solves this by consolidating data into an MCP server that any AI can query, turning fragmented metrics into coherent narratives that explain trends and anomalies.
The first major feature is freddy's ability to connect multiple sources via OAuth in a read-only, revocable, encrypted manner. Users can link Oura, WHOOP, Polar, Garmin, Dexcom, Wahoo, Hevy, Withings, Concept2, Intervals.icu, Suunto, and more with a single click, granting freddy access to over 100 metrics including HRV, glucose, sleep stages, and power output. This works because freddy is built as an MCP server, meaning it speaks the Model Context Protocol that AI agents understand natively. By pasting one MCP URL into Claude Desktop or ChatGPT, the AI gains the ability to query real-time and historical data. The benefit is elimination of manual data entry and dashboard hopping: the AI pulls exactly the numbers it needs, when it needs them, for a truly conversational analytics experience.
The second major feature is freddy's ability to perform cross-metric query and trend analysis. Once connected, users can ask questions like 'Why did I sleep like garbage last night?' and freddy responds with specific metric context: HRV dropped 18 ms below the 30-day baseline, resting HR up 7 bpm, with three wake events between 2 and 4 am. It can also detect trends, such as a 30-day HRV baseline drift from 56 to 49 ms alongside a 34% climb in training load, while total sleep remains unchanged but fragmentation is up 22%. This analysis is powered by freddy's access to data from 21 sources, enabling it to correlate sleep fragmentation, glucose spikes, and training fatigue in ways single-device apps cannot, providing answers that are both specific and contextual.
Publisher
A
admin
Launch Date2026-06-07
Pricingfree
Domain Authority
Ahrefs DR1/100