SuperAPI AI Directory
Partner with us中文AI newsFavorites
All toolsAI healthcarewellness-project-mcp

wellness-project-mcp

🎖️ ☁️ - Read workouts, sleep, nutrition and recovery data from Apple Health, Fitbit, Oura and Google Health Connect, with inline chart widgets.

Open website

About this tool

Purpose

Wellness Project MCP is a hosted MCP (Model Context Protocol) server for bringing connected health and fitness data into MCP-compatible assistants such as Claude, ChatGPT, Gemini Spark, Grok, and Mistral. After a user's devices and logs sync to the same account, supported assistants can read or update workouts, nutrition, sleep, recovery, body metrics, goals, labs, health status, and more using natural language.

This GitHub repository is public documentation and a directory, not a self-hosted backend, and it does not contain the private server implementation.

Main capabilities

  • Connect multiple AI assistants through one hosted endpoint.
  • Natural language Q&A, for example: "How did my training go this week?" "Show my sleep for the past month." "How is my bench press progressing?" "Log today's lunch: chicken burrito bowl, about 700 calories." "How does my resting heart rate trend this quarter compare with HRV?" "How am I doing on my goals?" "Log 16 ounces of water."
  • Read tools return recorded and synced health data; write tools add, update, or delete entries when the user asks.
  • Component tools can render structured interactive views in compatible MCP Apps UI clients instead of displaying large amounts of numbers.
  • The directory contains 73 public tools generated from the production tool registry, as well as 15 active interactive components.

How to use

  1. First, at https://wellnessproject.ai create an account, and connect the needed health data sources in Settings → Devices.
  2. All clients share the same hosted endpoint:https://wellnessproject.ai/api/mcp。认证由 OAuth is handled, and the current public connection flow does not use personal API keys.

Claude (web version)

  1. Open Claude on the web.
  2. Go to Customize → Connectors → Add custom connector.
  3. Paste https://wellnessproject.ai/api/mcp。
  4. If Claude shows Client ID and Client Secret fields, leave them blank. Claude can dynamically register on its own.
  5. Log in to Wellness Project and authorize the connection.

Claude Code

Claude Code installs Wellness Project as a plugin, and the plugin contains only the configuration for this server and a set of setup skills.

  1. Restart Claude Code.
  2. Run /mcp, select wellness-project, and authorize. No Client ID, Client Secret, or API key needs to be entered.
  3. If you need guidance, run /wellness-project:setup.

Cline

Cline connects to Wellness Project as a remote MCP server, without cloning, installing, or running anything locally.

  1. Open the MCP Servers icon in the Cline sidebar and select Remote Servers.
  2. Enter a server name, for example wellness-project.
  3. Paste https://wellnessproject.ai/api/mcp as the server URL and add it.
  4. When Cline opens it in the browser, complete the Wellness Project OAuth flow.

To edit cline_mcp_settings.json directly: explicitly set "type": "streamableHttp". If type is omitted, Cline falls back to the legacy SSE transport. No Client ID, Client Secret, or API key needs to be entered. llms-install.md contains the same steps for agents performing installation.

ChatGPT

ChatGPT custom MCP apps use developer mode. Current OpenAI availability depends on the plan: full MCP support including write/modify operations is available in beta for Business, Enterprise, and Edu; Pro can connect to MCP with read/fetch permissions in developer mode.

  1. Enable developer mode for the account or workspace in the way allowed by the ChatGPT plan.
  2. Create a custom MCP app from ChatGPT's Apps settings.
  3. Use https://wellnessproject.ai/api/mcp as the remote MCP server.
  4. Complete the Wellness Project OAuth flow.

OpenAI's current settings and plan details:https://help.openai.com/en/articles/12584461

Gemini Spark

Custom MCP apps require Gemini Spark access and are added from the Gemini web app.

  1. Open https://gemini.google.com。
  2. Open Settings & help → Connected Apps; if needed, first open Personal Intelligence → Connected Apps.
  3. Under Custom apps for Spark, select Add a custom app.
  4. Enter https://wellnessproject.ai/api/mcp and click Next.

Google's current eligibility and availability rules:https://support.google.com/gemini/answer/17209137

Grok

  1. Open https://grok.com/connectors，选择 New Connector → Custom.
  2. If Client ID and Client Secret fields appear, leave them blank.
  3. Grok connector availability may depend on the Grok plan.

Mistral Le Chat

  1. In Le Chat, open Intelligence → Connectors → Add Connector → Custom MCP Connector.
  2. Workspace permissions may require an owner or admin to add the connector.

Interactive Components

Fifteen active tools render interactive views through the MCP Apps UI extension in compatible clients. Many trend components support multiple time ranges; detail components use dates or records relevant to the request.

The complete list of 73 public tools (grouped by domain) is in TOOLS.md. Machine-readable descriptions, annotations, and JSON Schema are in catalog/tools.json. Both files are generated from the same production tool registry used by tools/list. Admin-only tools and retired tools invoked only by legacy cached clients are not shown in that catalog. Do not manually edit the generated catalog.

Data Sources and Integrations

Wellness Project combines native mobile health storage, direct integrations, intermediary sources, and manual/chat records. Garmin, Samsung Health, Amazfit/Zepp, Coros, Wahoo, Strava, Peloton, smart scales, and other compatible sources can also reach Wellness Project through them when they write relevant data to Apple Health or Health Connect.

Pricing and Status

The material states that prices reflect the current production configuration as of September 7, 2026, and local app store prices may differ. The material mentions a Free Basic free tier, and paid Founder Pro and standard Pro plans that add analytics capabilities and other paid features, and states that current prices are listed in the text, but does not provide specific price amounts; specific prices are not yet available.

Privacy and Authentication

  • The public MCP endpoint uses OAuth 2.1, and supported clients can register dynamically.
  • Every MCP request is scoped to the authenticated Wellness Project account.
  • The data layer uses row-level security to protect users' own records.
  • Connector access can be revoked without self-hosting or rotating long-lived personal API keys.
  • Wellness Project is informational software, not a medical product.

Use Cases and Limitations

Suitable for users who want to query, record, and manage personal health and fitness data in natural language within an AI assistant. Limitations include: the service is hosted, this repository does not include a server implementation, and it cannot be self-hosted; client-side MCP availability and permissions vary by product and plan; ChatGPT's MCP support depends on the plan; Gemini custom apps require Spark access; Grok connector availability may depend on the plan; Mistral workspace permissions may require an administrator. The material does not provide specific pricing, complete tool list details, or medical use claims; the relevant facts are not yet available.

Related tools

Browse category