GetFit
For assistants · MCP server

The workout that fits, asked for in a sentence.

Someone tells their assistant how much time they have, what gear is around, and what hurts. GetFit's coach builds the whole session — warm-up, sets, reps, rest, effort and every "stop if" — and hands it back in the chat. Free. No sign-in.

I want to work out legs tonight from home. I only have 30 minutes, and my knees are sore.
via GetFit · build_workout Here's a 30-minute bodyweight legs workout, easy on the knees: glute bridges, hamstring slides and a wall sit.
AGlute bridge3 × 15 · 45sStop if you feel it in your lower back.
ASupine hamstring slide3 × 12 · 45s
BSingle-leg Romanian deadlift3 × 8/side · 60sStop if your knee wobbles.
CWall sit3 × 40s · 45s
Run it in GetFit's Player →
Connect

One URL. No key.

Streamable HTTP, JSON-RPC over POST, stateless. GET returns 405. Add it as a custom connector in Claude, ChatGPT or any MCP client.

https://thegetfit.app/mcp-server
Transportstreamable-http
AuthNone (public tools)
Manifest/.well-known/mcp.json
PersonalMint a connector URL
Tools

Each one is a question.

ToolThe question it answers
build_workoutWhat workout can I do with this time, this gear, and this body? Inputs: focus, minutes, equipment, avoid, level, goal, intensity.
search_workoutsWhat workouts are on GetFit for legs / 20 minutes / dumbbells?
get_workoutShow me that workout in full.
list_contributorsWho posts workouts on GetFit?
get_contributorWho is this contributor and what have they posted?
my_profilepersonalWhat does GetFit know about me?
build_planpersonalMake me a 4-week plan, 3 days a week, and put it on my calendar.
my_schedulepersonalWhat's my workout today, or this week?
my_historypersonalHow has my training gone lately?
build_workoutpersonalThe same builder, using their profile; saves to their library, private unless they choose to post it.

Personal tools appear only on a person's own connector URL, minted on /connect. They also keep list, create, update, publish and delete for their own workouts.

How it behaves

Built for the chat.

Limits

Generous, and stated.

Policies

Plain terms.

What we log and for how long: Privacy. Something broken or a question: Support or [email protected]. Machine-readable: llms.txt · openapi.json.