# csjoblom/musclesworked-mcp [Health: Active]

**Category:** 🏃 Sports  
**Repository:** https://github.com/csjoblom/musclesworked-mcp  
**GitHub Stars:** 4  
**npm Downloads (last month):** 189  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/csjoblom-musclesworked-mcp

## Description
Exercise-to-muscle mapping MCP server. Look up muscles worked by 856 exercises across 65 muscles and 14 muscle groups, analyze workouts for gaps, and find alternative exercises ranked by muscle overlap.

## Tools
Capabilities this server exposes over MCP:

- **get_muscles_worked** — Get primary, secondary, and stabilizer muscles for an exercise
- **find_exercises** — Find exercises targeting a specific muscle with optional filters
- **analyze_workout** — Analyze a workout for coverage, gaps, and imbalances
- **get_alternatives** — Find alternative exercises ranked by muscle overlap
- **search_exercises** — Search exercises by name to discover IDs
- **search_muscles** — Search muscles by name to discover IDs

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "musclesworked-mcp": {
    "command": "npx",
    "args": ["-y","musclesworked-mcp"],
    "env": {
      "MUSCLESWORKED_API_KEY": ""
    }
  }
}
```

**Requires environment variables:** `MUSCLESWORKED_API_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What csjoblom/musclesworked-mcp MCP server does

The csjoblom/musclesworked-mcp MCP server exposes exercise and muscle mapping data from musclesworked.com to MCP-compatible clients. Its dataset covers 856 exercises, 65 muscles, and 14 muscle groups. Exercise results distinguish between primary, secondary, and stabilizer muscles, which makes the server suitable for questions that need more detail than an exercise-name search.

The server supports both individual exercise lookups and broader workout analysis. An agent can identify the muscles involved in an exercise, locate exercises for a selected muscle, inspect a workout for coverage gaps or imbalances, and find alternatives ranked by how much muscle activation they share with the original movement.

## How it works

The MCP client starts the local `musclesworked-mcp` package through `npx`. The process communicates with the musclesworked.com API and authenticates with an API key. The key can be supplied through the `MUSCLESWORKED_API_KEY` environment variable or through the server's `--api-key` option.

Two search tools help discover the identifiers needed for other queries. `search_exercises` searches exercise names, while `search_muscles` searches muscle names. Once an exercise or muscle ID is known, the remaining tools can retrieve mappings, filter exercise results, analyze a workout, or rank alternatives.

The base API URL defaults to `https://musclesworked.com`. It can be changed with `MUSCLESWORKED_API_URL`, which is useful when the API endpoint needs to be overridden. The provided setup examples show local stdio configuration for Claude Desktop, Claude Code, and Cursor.

## Tools and capabilities

The csjoblom/musclesworked-mcp MCP server exposes six tools:

- `get_muscles_worked`: Returns the primary, secondary, and stabilizer muscles associated with an exercise.
- `find_exercises`: Finds exercises that target a specified muscle, with optional filters.
- `analyze_workout`: Reviews a workout for muscle coverage, gaps, and imbalances.
- `get_alternatives`: Finds replacement exercises and ranks them by muscle overlap.
- `search_exercises`: Searches exercise names to help discover exercise IDs.
- `search_muscles`: Searches muscle names to help discover muscle IDs.

These capabilities support prompts such as identifying the muscles used by a barbell bench press, finding bodyweight exercises for the chest, reviewing a push-day exercise list, comparing squat alternatives, or searching for hamstring exercises. The server supplies exercise mapping data; the provided material does not describe coaching guidance, workout prescription, or medical assessment features.

## Setup and configuration

Create an API key through the musclesworked.com dashboard before starting the server. A Claude Desktop or Cursor configuration can run `npx -y musclesworked-mcp` and pass the key in the environment. Claude Code can use the same package and provide the key with `--api-key`.

For local development, the repository instructions use `npm install` followed by `npm run build`. The built server can be run from the project output, and the README shows MCP Inspector usage for testing. The project is licensed under MIT.

The csjoblom/musclesworked-mcp MCP server does not document a no-key mode. Keep the API key out of prompts and source-controlled configuration files when deploying it for regular use.

_Full upstream README: https://allmcps.com/mcp/csjoblom-musclesworked-mcp/readme_