mcpField
T
Tally Workout Tracker

Tally Workout Tracker

Remote
github.com/westveghAI & Agents0Updated 17w ago
Repository

About Tally Workout Tracker

Tally Workout Tracker is an iOS workout log for strength training, with a remote MCP server that lets Claude read training history and write plans back to the app. Changes made in the app are protected from being overwritten.

Key features

  1. 01Read training context and workout history
  2. 02Search the exercise library
  3. 03Create or replace a planned day
  4. 04Plan workouts across multiple days
  5. 05Protect edits made in the app
  6. 06OAuth 2.1 account access

Use cases

  1. 01Discuss plans using recent workout history
  2. 02Add a week of training to Tally
  3. 03Look up completed sets, reps, and weights
  4. 04Preserve in-app edits while updating plans

MCP support

Provides a Streamable HTTP remote MCP server that lets Claude and other clients read workout data and write plans using OAuth 2.1. Its tools include read and write operations, with protection for changes made in the app.

Configuration

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

Endpointhttps://mcp.tallyworkout.app/api/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

Tally MCP — the workout app Claude can write to

Tally is a fast, offline-first workout tracker for iOS. This is its first-party remote MCP server: Claude reads your real training history and writes planned workouts back into the app. Talk through your program in conversation — exercises, sets, progressions — say "put it in Tally," open the app, and your week is planned, synced, and ready to log.

Most fitness MCPs are read-only. Tally's is a full round-trip: plan conversationally, log at the gym, and Claude respects your gym-floor edits — anything you change in the app is treated as truth and never overwritten.

Server URL: https://mcp.tallyworkout.app/api/mcp Transport: Streamable HTTP · Auth: OAuth 2.1 (PKCE, dynamic client registration) Cost: free — bring your own Claude subscription

Requirements

Connect

claude.ai (one-time, in a browser)
  1. Copy the server URL: https://mcp.tallyworkout.app/api/mcp
  2. In a browser, open claude.ai → Settings → Connectors → Add custom connector. Name it Tally, paste the URL.
  3. Sign in with your Tally account when asked. Then try: "Plan my next week of training and add it to Tally."

Connectors can't be added from Claude's phone app itself; once added in a browser they sync to Claude's mobile and desktop apps automatically.

Claude Code
claude mcp add --transport http tally https://mcp.tallyworkout.app/api/mcp
Any MCP client (Streamable HTTP + OAuth 2.1)
{
  "mcpServers": {
    "tally": {
      "type": "http",
      "url": "https://mcp.tallyworkout.app/api/mcp"
    }
  }
}

Tools

ToolAccessWhat it does
get_training_contextreadTraining snapshot: recent sessions, streaks, patterns — grounds programming advice in what you actually did
get_workout_historyreadCompleted workout history (sets, reps, weights) over a date range
get_planned_workoutreadRead a planned day (for read-modify-write edits)
search_exercisesreadSearch Tally's exercise library (800+ movements)
plan_workout_daywriteCreate or replace one planned day
plan_workout_rangewritePlan a multi-day range — whole weeks or blocks in one call

All tools carry MCP annotations (readOnlyHint / destructiveHint). Writes never touch days you've edited in the app (user_modified protection), and every write is stamped with a batch id for auditability.

Privacy & security

  • OAuth 2.1 with PKCE; access is scoped to your Tally account and revocable anytime from Claude's connector settings.
  • Server-side analytics record tool names and timing only — never workout content.
  • Privacy policy: mcp.tallyworkout.app/privacy

About this repo

The server implementation lives in Tally's backend (closed source); this repo is the connector's public home — docs, config snippets, and the issue tracker for bug reports and tool requests.

Built by Vegh Labs LLC. Not affiliated with or endorsed by Anthropic; Claude is a trademark of Anthropic, PBC.

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.