mcpField
P
Prospre

Prospre

Remote
prospre.ioHealth0Updated 6d ago

About Prospre

Prospre is a plugin that connects an AI assistant to a Prospre meal-planning and macro-tracking account. It supports food logging, meal planning, diary corrections, grocery lists, and nutrition target adjustments through guided workflows.

Key features

  1. 01Log a day of food from a description
  2. 02Plan meals based on recorded intake
  3. 03Correct food diary entries
  4. 04Read and update grocery lists
  5. 05Adjust nutrition targets as weight changes

Use cases

  1. 01Logging a day's meals from a natural-language description
  2. 02Planning next week's meals from recent intake
  3. 03Correcting diary items and checking off groceries

MCP support

The plugin connects clients to Prospre's MCP server at `https://www.prospre.io/mcp`. Authentication uses Prospre OAuth, and the bundled skills guide clients through tool workflows.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://www.prospre.io/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

Prospre plugin

This plugin connects an AI assistant to a Prospre account. Prospre is a meal planning and macro tracking app. With the plugin, the assistant can log a whole day of food from a description, plan the coming week from what the user actually ate, fix diary entries after the fact, read and tick off the grocery list, and adjust nutrition targets as the user's weight moves. It works in Claude, ChatGPT, Grok Build, Claude Code and any other client that can load the files in this repository.

What is in the repository

  • .mcp.json points the assistant at Prospre's MCP server, https://www.prospre.io/mcp.
  • skills/ holds one skill per workflow. Each names the Prospre tools to call, in order, the defaults to apply, and what a good reply looks like. get-started explains the connection to a new user.
  • .claude-plugin/plugin.json is the plugin manifest.
  • chatgpt/ holds the manifest and build script for the ChatGPT package.
  • mcp-registry/server.json is the listing for the official MCP registry (docs/mcp-registry.md).
  • docs/ holds the submission notes for each directory.

The plugin is Markdown and JSON only. It runs no scripts, installs no packages and sets no hooks.

Using it

Claude (web, desktop and mobile): add Prospre from the plugin directory, then connect the bundled Prospre connector on the plugin's Connectors tab and sign in with a Prospre account.

Claude Code: claude --plugin-dir /path/to/prospre-plugin, then /mcp, choose prospre and authenticate. The skills appear as /prospre:log-day and so on.

Grok Build: grok --plugin-dir /path/to/prospre-plugin, or install it from the Grok Build marketplace once listed. Authentication starts in the browser on first use.

ChatGPT: install Prospre from the plugin directory once listed, or build the package with chatgpt/build.sh and add it through a local marketplace.

Every path needs a Prospre account. The free tier is enough. There is no API key: the assistant signs in through Prospre's OAuth consent screen, and the user approves exactly what it can reach.

What the plugin sends, and where

The assistant talks to one network endpoint, https://www.prospre.io/mcp, over HTTPS with the OAuth token Prospre issues at sign-in. The credential it needs is that token; the plugin stores nothing itself and asks for no other secret. When a skill runs a Prospre tool, the request carries what the user asked for, such as a food name, an amount, a date or a weight, and the result comes back to the assistant: diary entries, saved meals and recipes, the grocery list, weigh-ins, targets, and the name and email on the account. The assistant's own provider then handles that result under its privacy policy. Prospre does not receive the conversation, only the tool requests. Prospre's privacy policy is at https://www.prospre.io/privacy-policy, and the data each tool reads or changes is listed on https://www.prospre.io/mcp-server.

One tool, get_grocery_order_link, also sends the names and quantities on the grocery list to Instacart or Amazon Fresh to build an ordering link, and that link is affiliate-tagged. The skill that uses it tells the user both.

Trademark and license

Prospre is a trademark of Prospre. Use of the service is governed by https://www.prospre.io/terms-of-service. The files in this repository are released under the MIT license in LICENSE.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.