MCP · Included with FitMaxx Pro

Your fitness data.
Your kind of AI.

Log sets, explore your training history, and build a progress dashboard that’s entirely yours. Connect your AI assistant to FitMaxx.

Sign in with Apple or Google. Use your existing FitMaxx account.

FITMAXX / YOUR OWN VIEWIllustrative data
“Build me a dashboard of my last 90 days of training.”
Your work, adding up.LAST 90 DAYS
36Workouts
108Sets this month
3×Weekly rhythm
Bench press · working weight50 → 65 kg
Week 1Week 6Week 12

An example of what your agent can build. Your data. Your layout.

Connect once. Keep training.

Choose your assistant.

You’ll need an active FitMaxx Pro subscription and the Apple or Google account you use in the app.

SERVER URL
https://api.fitmaxxx.com/mcp

Codex

CLI setup

With the Codex CLI installed, run these commands in Terminal:

codex mcp add fitmaxx --url https://api.fitmaxxx.com/mcp

codex mcp login fitmaxx --scopes profile:read,workouts:read,workouts:write,body:read,body:write,training:read,data:export,offline_access

Complete sign-in in your browser, approve access, then start a new Codex session. Check the connection with codex mcp list.

After connecting, try a read first.
Use FitMaxx to show my profile.

Start with a question

More ways to use your work.

02 / Keep your log moving

Say the sets. Save the work.

Log exercises and body measurements from your assistant, then see them in FitMaxx.

Log three sets of barbell bench press at 50 kg, eight reps each.
03 / Find your pattern

Ask better questions.

Look across your history to understand how your training is changing.

Use FitMaxx to compare my bench press performance over the last eight weeks. Show working weights, reps and training frequency, and flag missing data.

Your account. Your choice.

You decide what’s connected.

FitMaxx MCP is included with every active Pro plan. Sign in using the same Apple or Google account you use in the app and review the permissions you approve.

Your assistant can access only the data and actions allowed by those permissions. Information requested through an assistant is shared with that assistant’s provider.

Manage connected apps

Disconnect to revoke future access. This does not erase copies already saved by your assistant or exported into a dashboard.

A few useful answers

Before your first set.

What is MCP?

Model Context Protocol is a standard that lets an AI assistant use external tools. FitMaxx provides those tools so your assistant can read your training data and perform actions you authorize.

I’m Pro in the app. Why can’t I connect?

Check that your subscription is active and that you signed in with the same provider and account as in FitMaxx. An Apple login and a Google login can belong to different accounts. Restore purchases in the app if needed, then retry. If access still fails, contact support@fitmaxxx.com.

Do I need a separate subscription?

There is no separate FitMaxx MCP subscription. You need active FitMaxx Pro access. Your AI assistant may have its own subscription or account requirements.

Which assistants have been tested?

Codex has passed checks for reads, writes and token refresh. Claude Code, Claude and ChatGPT guides are setup previews until their end-to-end checks are complete.

My connection stopped working. What now?

Check your Pro subscription and reconnect through your assistant’s MCP settings. In Codex, rerun codex mcp login fitmaxx. To switch accounts, revoke the old connection in Connected apps, then sign in again with the correct account.

Can I export all my data for a dashboard?

Your agent can fetch a paginated snapshot of supported fitness data. Exports are bounded to 100,000 rows and 100 MiB, and snapshots expire after 24 hours. Ask your agent to fetch every page and show any limits or missing data. A downloaded dashboard is a snapshot; refresh it to include later workouts.

Technical reference: transport, permissions and API

The endpoint uses Streamable HTTP with OAuth and PKCE. Read scopes: profile:read, workouts:read, body:read, training:read. Write scopes: workouts:write, body:write. Export uses data:export; ongoing access uses offline_access.

Tools cover profiles, workout sessions and sets, body check-ins, training history, progress and exports. Your client discovers the available tools after authentication. REST grants can download JSON or NDJSON exports; MCP grants use paginated tool results.

Public API schema (OpenAPI JSON) ↗