Run your gym from an AI assistant.
The Model Context Protocol is an open standard that lets assistants like Claude call tools. We ship an open-source MCP server that wraps the Gym Seva API — so you can ask questions in plain language and get answers from your live gym data.
renewals_dueEleven read-only tools, out of the box.
Every tool is a thin wrapper over an API endpoint. Nothing writes or deletes — the starter is read-only by design. Add write tools when you’re ready.
whoamiThe signed-in staff account and its roles.
gym_dashboardThe day’s headline numbers for the current gym.
list_membersSearch members by term and status.
get_memberOne member’s full profile — plan, payments, sessions.
renewals_duePlans expiring soon or lapsed, with revenue at risk.
accounts_receivableSubscriptions with an outstanding balance.
kpi_matrixMonth-wise sign-ups, revenue, renewals and sessions.
premium_membersMost valuable members by lifetime spend.
list_leadsThe sales pipeline, optionally filtered by status.
pending_sessionsSessions awaiting member PIN approval.
cash_on_handCurrent posted cash-drawer balance.
Three steps to connect it.
1 — Install
The server lives in mcp-server/ in your Gym Seva repo. It’s a standalone Node package — no build step, Node 18+.
2 — Add your credentials
The server signs in with a staff account to get a session cookie — the API has no API keys, so use a manager or owner login for full report access. Need scoped programmatic access instead? [email protected].
3 — Point Claude at it
Add this to claude_desktop_config.json and restart. The same block works for Claude Code and any MCP client.
Build your own tools on it.
The MCP server is a thin, readable wrapper over the same REST API the whole product runs on. Extend it, or build a fresh integration straight from the reference.