fitdays-mcp-server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@fitdays-mcp-server show me my latest weight and body composition from FitDays"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
fitdays-mcp-server
Model Context Protocol server that exposes the unofficial FitDays / iComon smart-scale API as tools that an LLM client (Claude Desktop, Claude Code, etc.) can call.
It logs into your FitDays account once, caches a full ~6-year sync in memory, and serves the data as structured tool responses.
Tools
Tool | What it does |
| Sub-users (people) registered under the account. |
| FitDays-compatible devices known to the account. |
| Body-composition / weight measurements. Filters: |
| Most recent weight measurement (filter by |
| Force-refresh the cached sync data. Returns counts including the |
Related MCP server: Renpho Health MCP
Install
npm install -g fitdays-mcp-serverRequires Node.js ≥ 22.
Configuration
The server reads credentials from environment variables:
Variable | Required | Default |
| yes | — |
| yes | — |
| no |
|
Use with Claude Desktop / Claude Code
Add it to your MCP client config (claude_desktop_config.json or
~/.claude.json / .mcp.json):
{
"mcpServers": {
"fitdays": {
"command": "npx",
"args": ["-y", "fitdays-mcp-server"],
"env": {
"FITDAYS_EMAIL": "you@example.com",
"FITDAYS_PASSWORD": "your-password"
}
}
}
}Run with Docker
A multi-arch image (linux/amd64, linux/arm64) is published to GitHub
Container Registry on every release, tagged with the release version
(1.2.0, 1.2, 1) and latest:
docker run -d --name fitdays-mcp \
-e FITDAYS_EMAIL=you@example.com \
-e FITDAYS_PASSWORD=your-password \
-p 8000:8000 \
ghcr.io/roquerodrigo/fitdays-mcp-server:latestThe container serves the server over
Streamable HTTP
at http://localhost:8000/mcp, with a health endpoint at /healthz, so
remote MCP clients can connect to it directly:
{
"mcpServers": {
"fitdays": {
"type": "http",
"url": "http://localhost:8000/mcp"
}
}
}Credentials are read only from the environment; nothing is baked into the image. To use the stdio transport instead, run the server binary as the container command:
{
"mcpServers": {
"fitdays": {
"command": "docker",
"args": [
"run", "-i", "--rm",
"-e", "FITDAYS_EMAIL", "-e", "FITDAYS_PASSWORD",
"ghcr.io/roquerodrigo/fitdays-mcp-server:latest",
"fitdays-mcp-server"
],
"env": {
"FITDAYS_EMAIL": "you@example.com",
"FITDAYS_PASSWORD": "your-password"
}
}
}
}Run from source
git clone https://github.com/roquerodrigo/fitdays-mcp-server.git
cd fitdays-mcp-server
npm install
npm run build
FITDAYS_EMAIL=… FITDAYS_PASSWORD=… npm startSupport
This MCP server is built and maintained on personal time, on hardware bought for the purpose. If it is useful to you, consider sponsoring the work — it keeps the devices, the testing and the releases coming.
License
MIT
Available Tools
6 toolsget_latest_weightLatest weightA
Return the most recent body-composition / weight measurement in the current global cache, optionally for a single sub-user (suid).
Data is fetched lazily when no valid cache exists and stored in a global cache shared by all tools for 5 minutes; subsequent queries use that cached snapshot until it expires. Use refresh_sync if you are within the 5-minute cache window and fresher data is required.
Returns exactly one measurement containing weight_kg, weight_lb, bmi, bfr_pct (body fat percentage), rom_pct (muscle percentage), rosm_pct (skeletal muscle percentage), vwc_pct (body water percentage), pp_pct (protein percentage), sfr_pct (subcutaneous fat percentage), uvi (visceral fat index), bm_kg (bone mass), bmr_kcal (basal metabolic rate), bodyage (body age), measured_at (ISO 8601 timestamp), measured_time (Unix-seconds timestamp), data_id, suid, uid, is_deleted, and ext_data, or null if no matching measurement exists.
ext_data contains the measured person's context and FitDays reference values for interpreting the weight metrics. Treat these as informational rather than authoritative, and adjust them to the person's actual situation.
Context fields describe the person and measurement state, including age, height, sex (female or male), and onlyMeasureWeight (whether the record contains weight only).
Range and standard fields provide reference targets: Min and Max define the reference range, while Standard is the recommended value. They include body-fat mass, body-fat percentage, BMI, basal metabolic rate, bone mass, muscle mass, skeletal muscle mass, protein mass, water mass, and weight fields.
Control fields indicate the suggested adjustment toward the corresponding Standard value, including bfmControl (body-fat mass), ffmControl (fat-free mass), and weightControl (weight).
Fat-free-mass reference fields provide the recommended fat-free mass, including ffmStandard.
Index fields provide skeletal-muscle indexing, including smi (skeletal muscle index).
Classification fields describe body-composition status, including bodyScore, bodyType, and obesityDegree.
Target fields describe target body-composition values, including targetWeight, targetBodyfatMass, and targetSMMMass; unavailable target-mass values are null.
Device fields identify the source device, including deviceModelExt, deviceNameExt, and deviceSoftwareVer.
By default ignores tombstoned records (is_deleted: 1).
Inputs: suid, include_deleted, include_ext_data.
get_todays_weigh_insToday's weigh-insA
Fetch all body-composition / weight measurements for a single calendar day (defaults to today in America/Chicago or FITDAYS_TIMEZONE) in one call.
Returns date, timezone, count, has_weigh_in (boolean), latest (the most recent weigh-in for that day with full ext_data, or null if none), and weigh_ins (all weigh-ins for that day ordered newest first).
By default forces a fresh sync (refresh: true) and excludes deleted records (include_deleted: false) while including full ext_data (include_ext_data: true).
ext_data contains the measured person's context and FitDays reference values for interpreting the weight metrics. Treat these as informational rather than authoritative, and adjust them to the person's actual situation.
Context fields describe the person and measurement state, including age, height, sex (female or male), and onlyMeasureWeight (whether the record contains weight only).
Range and standard fields provide reference targets: Min and Max define the reference range, while Standard is the recommended value. They include body-fat mass, body-fat percentage, BMI, basal metabolic rate, bone mass, muscle mass, skeletal muscle mass, protein mass, water mass, and weight fields.
Control fields indicate the suggested adjustment toward the corresponding Standard value, including bfmControl (body-fat mass), ffmControl (fat-free mass), and weightControl (weight).
Fat-free-mass reference fields provide the recommended fat-free mass, including ffmStandard.
Index fields provide skeletal-muscle indexing, including smi (skeletal muscle index).
Classification fields describe body-composition status, including bodyScore, bodyType, and obesityDegree.
Target fields describe target body-composition values, including targetWeight, targetBodyfatMass, and targetSMMMass; unavailable target-mass values are null.
Device fields identify the source device, including deviceModelExt, deviceNameExt, and deviceSoftwareVer.
Inputs: date, suid, refresh, timezone, +2 more.
get_weight_historyWeight historyA
Return body-composition / weight measurements, optionally filtered by sub-user (suid) and time window.
Data is fetched lazily when no valid cache exists and stored in a global cache shared by all tools for 5 minutes; subsequent queries use that cached snapshot until it expires. Use refresh_sync if you are within the 5-minute cache window and fresher data is required.
Returns a list ordered newest first. Each measurement contains weight_kg, weight_lb, bmi, bfr_pct (body fat percentage), rom_pct (muscle percentage), rosm_pct (skeletal muscle percentage), vwc_pct (body water percentage), pp_pct (protein percentage), sfr_pct (subcutaneous fat percentage), uvi (visceral fat index), bm_kg (bone mass), bmr_kcal (basal metabolic rate), bodyage (body age), measured_at (ISO 8601 timestamp), measured_time (Unix-seconds timestamp), data_id, suid, uid, and is_deleted. Set include_ext_data: true to also return ext_data for each record.
ext_data contains the measured person's context and FitDays reference values for interpreting the weight metrics. Treat these as informational rather than authoritative, and adjust them to the person's actual situation.
Context fields describe the person and measurement state, including age, height, sex (female or male), and onlyMeasureWeight (whether the record contains weight only).
Range and standard fields provide reference targets: Min and Max define the reference range, while Standard is the recommended value. They include body-fat mass, body-fat percentage, BMI, basal metabolic rate, bone mass, muscle mass, skeletal muscle mass, protein mass, water mass, and weight fields.
Control fields indicate the suggested adjustment toward the corresponding Standard value, including bfmControl (body-fat mass), ffmControl (fat-free mass), and weightControl (weight).
Fat-free-mass reference fields provide the recommended fat-free mass, including ffmStandard.
Index fields provide skeletal-muscle indexing, including smi (skeletal muscle index).
Classification fields describe body-composition status, including bodyScore, bodyType, and obesityDegree.
Target fields describe target body-composition values, including targetWeight, targetBodyfatMass, and targetSMMMass; unavailable target-mass values are null.
Device fields identify the source device, including deviceModelExt, deviceNameExt, and deviceSoftwareVer.
By default includes tombstoned records (is_deleted: 1); set include_deleted: false to hide them.
Inputs: suid, limit, since, until, +2 more.
list_devicesList FitDays devicesA
List the devices bound to the FitDays account.
Returns device_id (FitDays device identifier), name, model, mac (MAC address), and firmware_ver (firmware version) for each device.
Inputs: no named parameters.
list_usersList FitDays usersA
List the sub-users (people) registered under the FitDays account.
Each user has a unique and stable suid, used by other tools to query data for that user.
Returns a list where each entry contains nickname, sex (male or female, biological sex), birthday, height_cm, target_weight_kg (user-set target weight), suid, and uid (account uid).
Inputs: no named parameters.
refresh_syncRefresh sync cacheA
Force-refresh the global FitDays cache shared by all tools.
FitDays data is loaded lazily: the first tool call that needs synchronized account data fetches a fresh snapshot and stores it in the global cache for 5 minutes.
Returns synchronized record counts for devices, users, height_records, and weight_records (active, deleted, total).
Inputs: no named parameters.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 6 tools
The tools are mostly clearly separated by purpose: users, devices, history, today's weigh-ins, latest measurement, and cache refresh. The three weight-retrieval tools overlap in data type, but their descriptions make the distinctions between filtered history, single-day convenience, and single latest record reasonably clear.
Most names follow a consistent snake_case verb_noun pattern, such as list_users, get_weight_history, and refresh_sync. Minor deviations include get_todays_weigh_ins using 'weigh_ins' while other tools use 'weight', which is readable but slightly inconsistent.
Six tools is well scoped for a read-only FitDays data server. Each tool covers a distinct access pattern or cache operation, and none feels redundant enough to remove.
The surface covers users, devices, weight history, daily weigh-ins, latest measurement, and cache refresh. A minor gap is that refresh_sync references height_records, but no tool exposes height-record history, so that part of the domain is not directly accessible.
Maintenance
Related MCP Connectors
Every Canopy public API operation as a tool, run as you after a browser sign-in or on an API key.
Read wearables and lab health data — sleep, activity, workouts, timeseries, lab tests and orders.
Read your workouts, history, and stats; create and schedule new workouts. Writes are additive only.
Your Garmin, Oura, WHOOP, Hevy and 35 more fitness apps, queryable from your own AI.
Related MCP Servers
- FlicenseAqualityDmaintenanceFetches body composition data (weight, body fat percentage) from the Tanita Health Planet API and exposes it as MCP tools for AI agents like Claude Desktop.51-
- AlicenseNot gradedqualityDmaintenanceEnables AI agents to query and analyze Renpho smart scale data (weight, body fat, etc.) by syncing it from the Renpho app and exposing it through MCP tools for conversational exploration.1MIT
- AlicenseNot gradedqualityCmaintenanceEnables access to Renpho smart-scale body-composition data via the Renpho Health cloud API, providing metrics, trends, and diagnostics to Claude.MIT
- FlicenseBqualityBmaintenanceExposes client health data through a consolidated tool surface covering sleep, exercise, nutrition, cardio, fitness, cognitive, lifestyle, trends, notes, and chat history. Enables LLMs to retrieve time-series vitals, periodic assessments, unstructured text, and rolled-up summaries via standardized envelopes.11-