hevy-mcp, built by Swrm Io, is an MCP server that connects Hevy workout data to AI assistants, enabling conversational queries and automated routine actions. The tool exposes workout logs, routine management, an exercise template index, and body measurements via the Hevy API so models can analyze training volume and suggest plan edits. It secures API keys with environment variables and runs over MCP stdio. The target users are Hevy customers and developers building AI-powered personal trainers who want programmatic, queryable access to workout data.

Enables direct, chat-driven access and edits to a user's Hevy data

The tool turns the Hevy API into an MCP-backed data source, so AI assistants can fetch and act on workout information. Supported endpoints include paginated workout history and total counts, routine listing and updates, exercise templates, and body measurement entries. Typical outcomes are conversational queries about past sessions and saved routine edits, with actions issued by an assistant that call Hevy endpoints on the user's behalf.

Analysis usefulness reflects Hevy data quality and the chosen assistant's outputs

Generated analysis depends on two data sources: the user's recorded Hevy logs and the assistant producing the interpretation. Community feedback specifically highlights muscle-group imbalance checks and volume analysis as practical outputs. Those results provide usable signals for training decisions, but any prescriptive coaching recommendation should be verified against raw workout entries or a qualified coach, because the tool supplies data rather than independent validation.

Requires specific environment and account prerequisites to operate

Deployment demands Node.js and MCP tooling: Node.js v18 or higher is required, and an MCP-compatible client such as Claude Desktop must be present to use the stdio transport. A valid Hevy API key and a Hevy Pro account are necessary for API access. Installation options include npx invocation or cloning and building the repository with Node, which aligns it with developer-oriented setup workflows.

Fits developer workflows but data routing depends on client and assistant choices

The server is oriented toward developers building agent integrations, and the project is community-built rather than an official Hevy product. It is designed to run on any system supporting the Model Context Protocol via stdio, so how and where data leaves a host depends on the MCP client and assistant configuration. The server uses environment variables for API keys, which helps keep credentials out of source code during integration work.

A practical integration for developers and Hevy users who accept model-led analysis

The tool is a practical option for developers and Hevy customers who need conversational access to workout records and in-app routine edits; its value depends on the quality of recorded workouts and the assistant producing interpretations. Expect to use the tool as a data conduit and pair generated recommendations with manual review or expert oversight when applying training changes to a program.