Provides tools for interacting with Fitbit Web API, enabling access to health data such as profile, daily activity, sleep logs, heart rate, weight logs, activity timeseries, and a health snapshot.
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., "@personal-fitbit-mcp-server how was my sleep last night?"
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.
Fitbit Web API と連携する MCP (Model Context Protocol) サーバーです。 Claude Desktop や LM Studio などの AI アシスタントから、自然言語で自身の健康データにアクセス・分析できます。
ツール | 説明 |
| ユーザープロフィールと接続デバイスの情報 |
| 歩数・カロリー・距離・活動時間のサマリー |
| 睡眠時間・ステージ（deep/light/REM/wake）・効率 |
| 安静時心拍数・心拍ゾーン別時間 |
| 体重・BMI・体脂肪率の推移 |
| 活動量の時系列トレンド |
| 全データを一括取得した健康サマリー |
「今日の睡眠はどうだった？」
「今週の歩数の推移を見せて」
「最近心拍数が高い日はある？」
「先月の体重変化を教えて」
「今日の活動と睡眠をまとめて」Related MCP server: Oura MCP Server
Python 3.10 以上
uv パッケージマネージャー
Fitbit アカウント（Fitbit デバイスを使用中であること）
Fitbit 開発者アカウント（無料）
Fitbit 開発者ポータル にアクセスし、Fitbit アカウントでログイン
"Register an App" から新規アプリを登録（設定値は下記参照）
項目 | 設定値 |
OAuth 2.0 Application Type | Personal |
Callback URL |
|
Default Access Type | Read Only |
Application Website URL |
|
登録後に表示される Client ID と Client Secret を控えておく
git clone https://github.com/<your-username>/personal-fitbit-mcp-server.git
cd personal-fitbit-mcp-server
uv synccp .env.example .env.env を開いて Client ID と Client Secret を記入します。
FITBIT_CLIENT_ID=your_client_id_here
FITBIT_CLIENT_SECRET=your_client_secret_here⚠️
.envには認証情報が含まれます。.gitignoreで除外済みですが、公開リポジトリへのコミットには注意してください。
~/Library/Application Support/Claude/claude_desktop_config.json を編集します。
{
"mcpServers": {
"fitbit-health": {
"command": "/path/to/uv",
"args": [
"--directory", "/path/to/personal-fitbit-mcp-server",
"run", "personal-fitbit-mcp-server"
],
"env": {
"FITBIT_CLIENT_ID": "your_client_id",
"FITBIT_CLIENT_SECRET": "your_client_secret"
}
}
}
}Note:
uvのフルパスはwhich uvコマンドで確認できます（例:/Users/username/.local/bin/uv）。 Claude Desktop は通常のシェルの PATH を継承しないため、フルパスの指定が必要です。
mcp.json に以下を追加します。
{
"mcpServers": {
"fitbit-health": {
"command": "/path/to/uv",
"args": [
"--directory", "/path/to/personal-fitbit-mcp-server",
"run", "personal-fitbit-mcp-server"
],
"env": {
"FITBIT_CLIENT_ID": "your_client_id",
"FITBIT_CLIENT_SECRET": "your_client_secret"
}
}
}
}初めてツールを呼び出すと、OAuth 2.0 認証フローが自動的に開始されます。
サーバーの stderr に Fitbit 認証 URL が表示される
ブラウザでその URL を開いて Fitbit にログインし、アクセスを許可する
http://localhost:3000/callback にリダイレクトされ「Authorization successful!」と表示されたら完了
トークンが ~/.fitbit-health-mcp/tokens.json に保存される（以降は自動更新）
set -a && source .env && set +a
uv run mcp dev src/fitbit_health_mcp/server.py -e .ブラウザで Inspector UI が開き、各ツールを対話的にテストできます。
docs/design.md はこのプロジェクトの開発前の原案設計書です。
実装過程での仕様変更（アプリ名変更・パラメータ設定等）により、実際のコードと一部異なる箇所があります。
実装の仕様は本 README およびソースコードを参照してください。
src/fitbit_health_mcp/
├── server.py # FastMCP サーバー定義・ツール登録
├── auth.py # OAuth 2.0 PKCE 認証フロー
├── client.py # Fitbit API クライアント
├── formatters.py # レスポンス整形
├── config.py # 設定値
└── tools/
├── sleep.py
├── activity.py
├── heart_rate.py
├── weight.py
├── snapshot.py
└── profile.pyFastMCP — MCP サーバー実装
httpx — 非同期 HTTP クライアント
OAuth 2.0 Authorization Code Grant with PKCE（自前実装）
stdio トランスポート（Claude Desktop / LM Studio 両対応）
レート制限: Fitbit API は 150 リクエスト/時間/ユーザーの制限があります。get_health_snapshot は1回の呼び出しで4リクエストを消費します。
トークンの保存場所: ~/.fitbit-health-mcp/tokens.json（個人利用想定）
対応スコープ: activity / heartrate / sleep / weight / profile / settings
このプロジェクトは Claude (Anthropic) の支援を受けて開発されました。 設計・実装・デバッグのすべてのコードは作者がレビューおよびテスト済みです。
This project was developed with the assistance of Claude (Anthropic). All code has been reviewed and tested by the author.
This project is licensed under the MIT License.
Get activity trends over time. resource: steps|distance|calories|minutesVeryActive|minutesFairlyActive|minutesLightlyActive|minutesSedentary. period: 7d|30d|3m|6m|1y
| Name | Required | Description | Default |
|---|---|---|---|
| period | No | 7d | |
| resource | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden. It discloses the accepted values for resource and period, which is useful behavioral context. However, it does not describe the return format, pagination, authentication needs, or any side effects. The output schema likely covers returns, but other behavioral aspects remain opaque.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is extremely compact, with the primary purpose front-loaded in a single clear sentence followed by concise parameter options. Every word earns its place, and there is no redundant information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple 2-parameter tool with an output schema, the description covers the essential aspects: purpose and parameter constraints. It does not clarify the distinction from get_daily_activity, but the name and description imply a time-series focus. Given the simplicity, it is reasonably complete, though it could benefit from a brief note on typical use cases or output shape.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must compensate. It explicitly enumerates valid values for both resource (steps|distance|... ) and period (7d|30d|... ), providing crucial constraints not in the schema. It does not explain the meaning of each value, but the lists themselves are sufficient for correct parameter selection in most cases.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Get activity trends over time' which identifies the specific verb and resource. The resource and period value lists further define the scope, distinguishing it from sibling tools like get_daily_activity or get_heart_rate.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives clear context for when to use this tool: for activity trends over time. It does not explicitly mention alternatives or exclusions, but the phrase 'over time' implies longitudinal data, which differentiates it from daily snapshots. No direct comparison to siblings, so it falls short of a 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get daily activity summary. Returns steps, calories, distance, active minutes.
| Name | Required | Description | Default |
|---|---|---|---|
| date | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden, but it only lists return fields. It does not disclose behavior such as default date handling, timezone considerations, or any limitations. The existence of an output schema reduces the need to explain return values, but the description lacks behavioral context.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single sentence that is front-loaded and free of fluff. It conveys the core purpose efficiently, though it under-specifies details.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the simplicity of the tool (one optional parameter, output schema present), the description still leaves critical gaps: no mention of the date parameter, no distinction from timeseries, and no guidance on expected input. The sentence is too terse for an agent to invoke correctly without additional assumptions.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0% and the description does not mention the 'date' parameter at all. The agent receives no guidance on the accepted date format, the meaning of null, or how the parameter affects results. With one parameter and zero coverage, the description must compensate but fails to do so.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's function with a specific verb ('Get') and resource ('daily activity summary'), and enumerates the return fields (steps, calories, distance, active minutes). The term 'summary' distinguishes it from the sibling 'get_activity_timeseries', which implies time-series data.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to use this tool versus alternatives like get_activity_timeseries. The description does not mention any prerequisites, exclusions, or specific contexts where this summary is preferred.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get today's comprehensive health summary. Combines activity, sleep, heart rate, and latest weight.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must carry the full burden of behavioral disclosure. It only states what data is combined, not any behavioral traits such as read-only nature, potential data latency, or whether the 'latest weight' might be from a previous day. There is no context on how the data is sourced or any limitations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, focused sentence that conveys the tool's purpose and scope without any unnecessary verbosity. Every word earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a zero-parameter tool with an output schema, the description sufficiently explains what the snapshot contains and its scope. It could be more explicit about the read-only nature, but given the simplicity and the presence of an output schema, it is adequate.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema has zero parameters, so the baseline is 4. The description adds the 'today's' timeframe, clarifying the temporal scope of the snapshot, which is useful semantic context beyond the empty schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's function: 'Get today's comprehensive health summary' with a specific verb and resource. It differentiates from sibling tools by listing the combined data types (activity, sleep, heart rate, weight), making it distinct from the single-metric getters.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage for obtaining an overall health summary but does not explicitly say when to use this instead of the individual getter tools. No exclusions or alternatives are mentioned, though the phrase 'comprehensive health summary' hints at its role as an aggregator.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get heart rate data. Returns resting HR and time in zones.
| Name | Required | Description | Default |
|---|---|---|---|
| date | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It does not mention any side effects, permissions, rate limits, data retention, or return format beyond the vague 'resting HR and time in zones.' The read-only nature is implied by 'Get' but not explicitly guaranteed.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise and front-loaded, consisting of two short sentences that immediately convey the tool's purpose and key outputs. There is no redundant information or filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Despite having an output schema, the tool description is incomplete. It lacks context about the date parameter, any limitations or constraints, and how the returned data might be filtered or formatted. For a tool with a single optional parameter, the description fails to integrate that parameter into the overall usage context.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema has one parameter (date) with zero coverage, and the description does not mention this parameter at all. It fails to explain how date affects the data returned, whether it's required for a specific range, or what the default behavior is. This lack of compensation for low schema coverage is a serious gap.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose with a specific verb and resource ('Get heart rate data') and further defines the scope by mentioning the specific metrics returned ('resting HR and time in zones'). This distinguishes it from sibling tools focused on sleep, weight, or activity data.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to use this tool versus alternatives such as get_daily_activity or get_health_snapshot. There are no usage scenarios, prerequisites, or exclusions mentioned.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get user profile and device info.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It only says 'Get', which implies a read operation, but does not disclose any behavioral traits such as authentication requirements or whether it returns all devices. The output schema mitigates some ambiguity, but the description adds minimal transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single concise sentence that states exactly what the tool does with no unnecessary words or repetition.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a no-parameter read tool with an output schema, the description is sufficient. It could be slightly enhanced by adding what kind of profile/device info is returned, but the output schema likely covers that.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has zero parameters and the schema coverage is 100%, so the description does not need to explain parameters. The baseline for zero-parameter tools is 4.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Get') and resource ('user profile and device info'), making the tool's purpose immediately clear. It distinguishes itself from sibling tools that focus on sleep, activity, heart rate, etc., by targeting profile/device data.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
There is no explicit guidance on when to use this tool versus alternatives, but the distinct resource (profile/device) implies its usage. A clearer statement about when to prefer this over sibling health-data tools would improve the score.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get sleep data for a date. Returns duration, stages, efficiency.
| Name | Required | Description | Default |
|---|---|---|---|
| date | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description implies a read-only operation ('Get') and specifies what is returned, which is helpful. Given that no annotations are provided, it does not disclose behaviors such as date formatting expectations, null date handling, or any potential errors, leaving some gaps.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two concise sentences deliver the essential purpose and return fields without any unnecessary words. Every word earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple one-parameter tool with an output schema, the description covers the core purpose and return fields. It lacks details about date formatting and null behavior, but the output schema likely handles return structure. Overall adequate but not exhaustive.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema shows a single 'date' parameter with type string/null and a default of null, but no description. The tool description mentions 'for a date' which aligns with the parameter but does not specify the expected format or semantics of null, so it adds minimal meaning beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool retrieves sleep data for a specific date and lists the returned metrics (duration, stages, efficiency). The resource is unambiguous and distinct from sibling tools like heart rate or weight logs.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The phrase 'for a date' provides clear context for when to use this tool—when sleep data for a specific date is needed. However, it does not explicitly mention alternatives or exclusions, which would elevate it to a 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get weight data over a period. Returns weight, BMI, fat%. period: 7d|30d|3m|6m|1y
| Name | Required | Description | Default |
|---|---|---|---|
| period | No | 30d |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It discloses the return fields and the allowed period values, implying a read-only operation. However, it does not mention rate limits, auth requirements, or error behavior.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single sentence followed by a compact parameter list. Every word contributes value, with no redundant details.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a one-parameter, read-only tool, the description covers the resource, return metrics, and parameter options. The presence of an output schema further specifies the return structure, making the description sufficiently complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema offers only a default value for 'period' with no description. The description compensates fully by listing the accepted values ('7d|30d|3m|6m|1y'), which is essential for correct invocation.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Get weight data over a period' and lists the specific return metrics (weight, BMI, fat%). This unambiguously distinguishes it from sibling tools like get_heart_rate or get_sleep_log.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description does not provide explicit guidance on when to use this tool versus alternatives. The purpose is clear enough to imply usage, but there are no named alternatives or exclusion criteria.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0get_activity_timeseriesget_daily_activityget_health_snapshotget_heart_rateget_profileget_sleep_logget_weight_logScored across 7 tools
Each tool targets a distinct resource: sleep, daily activity, heart rate, weight, activity trends, profile, and a combined snapshot. The health snapshot is explicitly a summary and does not overlap with the detailed metrics, so there is no ambiguity.
All tool names follow a consistent 'get_' prefix followed by a snake_case resource name (e.g., get_sleep_log, get_health_snapshot). This pattern is uniform and highly predictable.
With 7 tools, the server is well-scoped for personal Fitbit data retrieval, covering all major data categories without unnecessary redundancy. The count falls comfortably within the ideal range.
The server covers core health data (activity, sleep, heart rate, weight, profile) and provides a daily snapshot. Minor gaps include lack of heart rate trends over time and nutrition data, but these are not critical for a read-only personal health server.
Connect your health, fitness, nutrition, sleep, and wearable data to your AI assistant.
Access approved FITsociety clients, schedules, bookings and coaching data with AI assistants.
Connect your wearables, rings and training apps, then ask your AI about your own health data.
Garmin data in Claude & ChatGPT via the Garmin Health API. OAuth sign-in, no password sharing.
