Fitbit Fitness Insights
Get AI-powered insights from your Fitbit data. Query your fitness metrics, analyze trends, and ask questions about your activity.
Features
- - 📊 Daily activity summaries (steps, calories, distance, active minutes)
- 💓 Heart rate data and zones
- 😴 Sleep tracking and analysis
- 🏃 Workout/activity logs
- 📈 Weekly and trend analysis
- 🤖 AI-powered insights and Q&A
Prerequisites
Requires: Fitbit OAuth access token
Setup steps in INLINECODE0
Commands
Get Profile
CODEBLOCK0
Daily Activity
CODEBLOCK1
Returns: steps, distance, calories, active minutes (very/fairly/lightly/sedentary), floors
Steps Range
CODEBLOCK2
Example:
CODEBLOCK3
Returns: total steps, average steps, daily breakdown
Heart Rate
CODEBLOCK4
Returns: resting heart rate, heart rate zones with minutes in each zone
Sleep Data
CODEBLOCK5
Returns: duration, efficiency, start/end times, sleep stages
Logged Activities
CODEBLOCK6
Returns: workouts/activities logged (name, duration, calories, distance)
Weekly Summary
CODEBLOCK7
Returns: 7-day summary of steps and key metrics
AI Insights Usage
When user asks fitness questions, use the API to fetch relevant data, then provide insights:
Example queries:
- - "How did I sleep last night?" → fetch sleep data, analyze quality
- "Did I hit my step goal this week?" → fetch weekly summary, compare to goals
- "What was my average heart rate during workouts?" → fetch heart + activities, analyze
- "Am I more active on weekdays or weekends?" → fetch range data, compare patterns
Analysis approach:
- 1. Identify what data is needed
- Fetch via appropriate API command
- Analyze the data
- Provide insights in conversational format
Example Responses
User: "How did I do this week?"
Agent:
- 1. Fetch weekly summary
- Fetch recent sleep data
- Respond: "You had a solid week! Averaged 8,234 steps/day (up 12% from last week). Hit your 10k step goal 4 out of 7 days. Sleep averaged 7.2 hours with 85% efficiency. CrossFit sessions on Mon/Wed/Fri looking consistent!"
User: "Did I exercise today?"
Agent:
- 1. Fetch daily activities
- Fetch daily activity summary (active minutes)
- Respond: "Yes! You logged a CrossFit session this morning (45 min, 312 calories). Plus 28 very active minutes total for the day."
Data Insights to Look For
- - Trends: Week-over-week changes, consistency patterns
- Goals: Compare to 10k steps, exercise frequency, sleep targets
- Correlations: Sleep quality vs activity, rest days vs performance
- Anomalies: Unusual spikes or drops
- Achievements: Personal bests, streaks, milestones
Token Management
The skill automatically loads tokens from /root/clawd/fitbit-config.json and refreshes them when expired (every 8 hours).
Auto-refresh: Tokens are refreshed automatically - no manual intervention needed!
Manual refresh (if needed):
CODEBLOCK8
Override with environment variable:
CODEBLOCK9
Error Handling
- - Missing token: Prompt user to set FITBITACCESSTOKEN
- API errors: Check token validity, may need refresh
- No data: Some days may have no logged activities or missing metrics
See references/fitbit-oauth-setup.md for token management.
Fitbit 健身洞察
从您的 Fitbit 数据获取 AI 驱动的洞察。查询您的健身指标，分析趋势，并就您的活动提出问题。
功能特点
- - 📊 每日活动摘要（步数、卡路里、距离、活跃分钟数）
- 💓 心率数据及区间
- 😴 睡眠追踪与分析
- 🏃 锻炼/活动记录
- 📈 周度与趋势分析
- 🤖 AI 驱动的洞察与问答
前置条件
需要： Fitbit OAuth 访问令牌
设置步骤详见 references/fitbit-oauth-setup.md
命令
获取个人资料
bash
FITBIT
ACCESSTOKEN=... python3 scripts/fitbit_api.py profile
每日活动
bash
python3 scripts/fitbit_api.py daily [日期]
示例：
python3 scripts/fitbit_api.py daily # 今天
python3 scripts/fitbit_api.py daily 2026-02-08 # 指定日期
返回：步数、距离、卡路里、活跃分钟数（高强度/中等强度/轻度/久坐）、楼层数
步数范围
bash
python3 scripts/fitbit_api.py steps <开始日期> <结束日期>
示例：
bash
python3 scripts/fitbit_api.py steps 2026-02-01 2026-02-07
返回：总步数、平均步数、每日明细
心率
bash
python3 scripts/fitbit_api.py heart [日期]
返回：静息心率、各心率区间及对应分钟数
睡眠数据
bash
python3 scripts/fitbit_api.py sleep [日期]
返回：时长、效率、开始/结束时间、睡眠阶段
已记录活动
bash
python3 scripts/fitbit_api.py activities [日期]
返回：已记录的锻炼/活动（名称、时长、卡路里、距离）
周度摘要
bash
python3 scripts/fitbit_api.py weekly
返回：7天步数及关键指标摘要
AI 洞察使用说明
当用户提出健身相关问题时，使用 API 获取相关数据，然后提供洞察：
示例查询：
- - 我昨晚睡得怎么样？ → 获取睡眠数据，分析质量
- 我这周达到步数目标了吗？ → 获取周度摘要，与目标对比
- 我锻炼时的平均心率是多少？ → 获取心率+活动数据，进行分析
- 我在工作日还是周末更活跃？ → 获取范围数据，对比模式
分析方法：
- 1. 确定需要哪些数据
- 通过相应的 API 命令获取
- 分析数据
- 以对话形式提供洞察
示例回复
用户： 我这周表现如何？
助手：
- 1. 获取周度摘要
- 获取近期睡眠数据
- 回复：你这周表现不错！平均每天 8,234 步（比上周增长 12%）。7 天中有 4 天达到 10,000 步目标。平均睡眠 7.2 小时，效率 85%。周一/周三/周五的 CrossFit 训练看起来很有规律！
用户： 我今天锻炼了吗？
助手：
- 1. 获取每日活动
- 获取每日活动摘要（活跃分钟数）
- 回复：是的！你今天早上记录了一次 CrossFit 训练（45 分钟，312 卡路里）。此外全天共有 28 分钟的高强度活跃时间。
需要关注的数据洞察
- - 趋势： 周与周之间的变化、一致性模式
- 目标： 与 10,000 步目标、锻炼频率、睡眠目标的对比
- 相关性： 睡眠质量与活动量、休息日与表现
- 异常： 不寻常的峰值或下降
- 成就： 个人最佳、连续记录、里程碑
令牌管理
该技能会自动从 /root/clawd/fitbit-config.json 加载令牌，并在过期时自动刷新（每 8 小时）。
自动刷新： 令牌会自动刷新——无需手动干预！
手动刷新（如需）：
bash
python3 scripts/refresh_token.py force
使用环境变量覆盖：
bash
export FITBITACCESSTOKEN=manual_token
错误处理
- - 缺少令牌： 提示用户设置 FITBITACCESSTOKEN
- API 错误： 检查令牌有效性，可能需要刷新
- 无数据： 某些天可能没有记录活动或缺少指标
令牌管理详见 references/fitbit-oauth-setup.md。
