Description

运动健康管家 Pro — 专为 Open Claw + 飞书生态打造的 AI 健身管家。双模式打卡（事后总结 / 边练边记）、3 层 Session 防丢失、飞书云文档同步、极致 Token 优化。Route to this skill when the user mentions workouts, logs...

README (SKILL.md)

运动健康管家 Pro (Fitness & Workout Manager Pro)

Name: Fitness & Workout Manager
Author: ph13917403910

⚠️ 平台要求：本 Skill 依赖「飞书 (Feishu)」生态，需为 Open Claw 绑定的机器人开通消息读写与云文档管理权限。

专为 Open Claw 打造的 AI 健身管家。无论是健身房力量训练还是网球对战，通过自然语言即可完美记录——你最省心的赛博教练。

✨ 核心亮点

双模式无缝打卡 — 事后一句话总结（"打网球 2 小时感觉很累"）或沉浸式边练边记（"卧推 60kg 8个 3组"），AI 自动解析入库。
防遗忘 Session 保护 — 3 层生命周期管理（主动关怀 → 心跳巡检 → 硬截止），练完忘了说"结束"也能自动闭环，绝不丢失数据。
飞书全景同步 — 计划 & 周/月度统计自动推送至飞书云文档，内置 Cron 定时提醒。
极致 Token 优化 — 日常查询仅返回紧凑摘要，全量记录自动导出为 Markdown，拒绝上下文爆炸。

🤖 Agent 路由指南

Capabilities: Stateful session handling (fitness-log start/end), NLP workout parsing, localized JSON storage, Feishu document sync, auto-stale-session cleanup.
Trigger: Route here when the user mentions workout routines, logs exercises, asks for fitness stats (fitness-status), or provides unstructured post-workout feedback.

Quick Start

# Create a fitness plan
fitness-plan create --goals "muscle_gain,cardio" --sport_pref "gym,tennis" --frequency 4

# One-shot log (after workout)
fitness-log "gym: bench press 60kg x 8 x 3, squat 80kg x 5 x 4, 1.5 hours"

# Live session (during workout — log each exercise as you go)
fitness-log start gym
fitness-log "bench press 60kg x 8 x 3"
fitness-log "squat 80kg x 5 x 4"
fitness-log "felt great"
fitness-log end "good session overall"

# View status
fitness-status --week
fitness-status --month
fitness-status --export

Commands

fitness-plan

Command	Description
`fitness-plan init`	Interactive multi-turn Q&A to generate a plan
`fitness-plan create --goals ... --frequency N`	Non-interactive plan creation
`fitness-plan show`	Display the current plan (Markdown)
`fitness-plan update --frequency 4 --level advanced`	Modify an existing plan
`fitness-plan sync`	Push plan to Feishu doc

fitness-log

Two modes: one-shot (after workout) and live session (during workout).

Command	Description
`fitness-log "\x3Ctext>"`	One-shot: parse and log a single workout entry
`fitness-log start [type]`	Start a live session (e.g. `start gym`)
`fitness-log "\x3Cexercise text>"`	Append exercise to active session
`fitness-log end ["notes"]`	End session, auto-calculate duration, commit to log
`fitness-log status`	Show active session info (includes idle/stale warnings)

Live session workflow: The user sends messages during training. Each message is appended to the same session entry. When the user says "end", all exercises are merged into one log entry with auto-calculated duration (from start to end time).

fitness-status

Command	Description
`fitness-status`	Last 7 entries (compact, token-efficient)
`fitness-status --week`	Weekly summary with stats
`fitness-status --month`	Monthly summary (last 30 days)
`fitness-status --days N`	Custom period summary
`fitness-status --all`	All entries (capped at 50, with truncation notice)
`fitness-status --export [N]`	Export last N days (default 30) as Markdown file

Session Lifecycle & Auto-Close

Sessions that are not explicitly ended are handled automatically:

Condition	Behavior
Idle ≥ 30 min	Claw proactively asks the user if training is done (via AGENTS.md rule). At most once per conversation to avoid nagging.
Idle > 4 hours or next day	Session is marked stale and auto-closed on the next fitness-related action. Duration is calculated up to the last message (not current time).
Heartbeat detects stale session	HEARTBEAT.md rule runs `fitness-log end` to force-close, and Claw notifies user on next interaction.
New `start` while stale exists	Old stale session is auto-closed first, then new one begins.

This ensures no workout data is ever lost, even if the user forgets to say "end".

Token Efficiency

Default queries return compact summaries (7 entries or period stats) to minimize token usage.
--all is capped at 50 entries with a truncation notice.
For full history, use --export to generate a Markdown file that can be sent to the user directly, avoiding large inline output.

Data Storage

Plan: ~/.openclaw/workspace/fitness-skill/plan.json
Log: ~/.openclaw/workspace/fitness-skill/log.json
Active session: ~/.openclaw/workspace/fitness-skill/active_session.json (auto-deleted on end)
Exports: ~/.openclaw/workspace/fitness-skill/exports/

Feishu Integration

Set FEISHU_FITNESS_DOC_ID to enable plan/log sync to a Feishu doc.
sync_plan_to_feishu() — push plan Markdown
sync_log_to_feishu(days=30) — push log report Markdown
Daily reminders via scheduler.add_cron_job() at the user's configured time.

Usage Guidance

This skill appears to do what it says: local workout parsing, session handling, and optional Feishu sync. Before installing, consider: 1) Feishu integration is optional — only set FEISHU_FITNESS_DOC_ID if you want sync; ensure the Feishu bot credentials/permissions you provide are limited to the needed doc read/write and messaging scopes. 2) The skill stores all plans/logs under ~/.openclaw/workspace/fitness-skill — review or back up that directory if you care about privacy. 3) SKILL.md mentions runtime hooks (AGENTS.md, HEARTBEAT.md, scheduler/feishu_api) that are provided by the OpenClaw runtime; verify your runtime supplies those APIs or the sync/auto-close features may not work. 4) If you plan to enable Feishu sync or run this on a shared environment, inspect the skill's sync_to_feishu and scheduler code paths (present in the full fitness_skill.py) to confirm no additional network endpoints or credentials are used beyond the Feishu doc ID. 5) As always, run untrusted code in a constrained environment if you are unsure.

Capability Analysis

Type: OpenClaw Skill Name: fitness-skill Version: 1.2.2 The fitness-skill bundle is a legitimate workout management tool designed for the OpenClaw platform. It implements workout logging, session management, and Feishu integration using standard Python libraries and regex-based NLP parsing. Analysis of scripts/fitness_skill.py and the CLI wrappers shows that data storage is confined to the skill's local workspace, and network activity is limited to the documented Feishu document synchronization. No evidence of malicious intent, data exfiltration, or unauthorized command execution was found.

Capability Assessment

✓ Purpose & Capability

Name/description, manifest entry points, and included scripts align: the package ships CLI entry points (fitness-plan, fitness-log, fitness-status) implemented in the provided Python files and stores data under ~/.openclaw/workspace/fitness-skill. The only required binary is python3, which is appropriate for the delivered Python scripts.

ℹ Instruction Scope

SKILL.md instructs the agent to route workout and logging requests here and documents local JSON storage, session lifecycle, and optional Feishu sync. The instructions do not ask for unrelated credentials or to read unrelated system paths. Note: SKILL.md references AGENTS.md and HEARTBEAT.md rules and scheduler behavior (cron/heartbeat) that are not included in the bundle; those behaviors depend on the OpenClaw runtime and may be non-functional or require runtime-provided hooks.

✓ Install Mechanism

No install spec; the skill is effectively delivered as scripts and docs. requirements.txt declares no external packages and the code uses only Python standard library, so nothing external is downloaded or extracted during install.

✓ Credentials

No required environment variables. One optional env var (FEISHU_FITNESS_DOC_ID) is declared for Feishu document sync and is proportionate to the described Feishu integration. No other credentials or unrelated env vars are requested.

✓ Persistence & Privilege

The skill does persist data locally under the user's home (~/.openclaw/workspace/fitness-skill) which matches its purpose. always:false and normal autonomous invocation settings are used. The skill does not request system-wide changes or modify other skills' configs in the shipped code.

Version History

v1.2.2

- Rebranded skill as "运动健康管家 Pro" with a new, more engaging Chinese description and highlights. - Improved documentation with clear feature highlights, platform requirements, and agent routing instructions. - Refined trigger guidelines for more accurate skill invocation based on user intent. - No functional or code changes; all updates are documentation-only.

v1.0.1

- Added standalone scripts: fitness-log.py, fitness-plan.py, fitness-status.py. - Enables modular command execution and easier script management. - No changes to usage or workflows—functionality remains the same.

v1.2.0

No file changes detected for version 1.2.0; documentation and skill details remain unchanged. - No updates or modifications in functionality or commands. - All features and integrations are consistent with the previous version.

v1.0.0

平台要求与使用前置准备 (Important Prerequisites): 注意：当前 Skill 仅支持与「飞书」生态集成。安装使用前，请确保您的 Open Claw 已绑定飞书机器人，并在飞书开发者后台为该机器人开通以下权限：消息与群组: 接收单聊消息、获取用户发给机器人的单聊消息、发送单聊消息云文档 (Drive): 查看、评论、编辑和管理云空间中所有文件 (用于创建和更新健身计划与打卡文档) 写给人类用户: 告别繁琐的表格和健忘的脑子，这是为你量身打造的终极 AI 健身管家。不论是挥汗如雨的力量训练，还是周末两小时的网球，它都能精准捕捉，完美记录。双模式记录，随心所欲：事后一句话总结： “今天打了 2 小时网球，感觉很累”，AI 自动提取时长、运动类型和感受并入库。沉浸式边练边记：开启 fitness-log start，每做完一组发一条消息（如“卧推 60kg 8个 3组”），结束时自动汇总。防遗忘 Session 保护：练完忘了说“结束”？没关系！Claw 会在空闲时主动关怀提醒，或通过底层心跳检测自动为你闭环并保存记录，绝不丢失任何一次流汗的证明。全局掌控与飞书同步：随时提取周/月度统计摘要，自动将最新的训练报告和健身计划同步到你的飞书文档，并支持定时提醒，做你最严苛也最贴心的赛博教练。极致省流：专为大模型上下文优化的 Token 效率设计，日常查询只看紧凑摘要，全量数据自动转存为 Markdown 报告导出。 Platform Requirements & Prerequisites: Note: This Skill currently ONLY supports the Feishu (Lark) ecosystem. Before using, ensure your Open Claw is connected to a Feishu Bot with the following permissions enabled in the Developer Console: Messages: Receive private messages, Read private messages sent to the bot, Send private messages Docs/Drive: View, comment, edit, and manage all files in Drive (Required for syncing plans and logs to Feishu Docs) For Humans: Say goodbye to tedious spreadsheets. This is the ultimate AI fitness concierge built for your Open Claw. Whether you're lifting heavy at the gym or playing an intense 2-hour tennis match, it tracks everything effortlessly. Dual Logging Modes: One-Shot NLP: Just say "Played tennis for 2 hours, felt exhausted." The AI parses the type, duration, and feeling instantly. Real-Time Session: Say fitness-log start, log your sets on the go ("Bench press 60kg x 8 x 3"), and it compiles everything when you finish. Bulletproof Session Recovery: Forgot to say "end" after a workout? No problem. The Skill features a 3-layer protection system (Active Care, Heartbeat Timeout, Hard Cutoff) to auto-save your session. You'll never lose a workout record again. Insights & Feishu Sync: Get weekly/monthly summaries instantly. Automatically sync your fitness plans and 30-day logs to Feishu Docs, complete with cron-job reminders to keep you accountable. Token Efficient: Designed to save your API costs. Daily queries return compact summaries, while full histories are exported cleanly as Markdown files without bloating the chat context. For AI Agents: System Capabilities: Advanced fitness lifecycle management featuring state-machine session handling, dual-mode NLP logging, token-efficient localized JSON storage, and seamless Feishu document sync. Core Commands Routing: fitness-plan: Handle init (multi-turn Q&A), create, update, show, and sync. fitness-log: Handle one-shot logging (NLP parse) or stateful sessions (start, real-time append, end, status). Include auto-recovery for stale sessions. fitness-status: Handle history queries with --week, --month, --days, and --export to avoid token bloat. Trigger Conditions: Route to this skill when the user mentions workout routines, logs exercises (e.g., "squat 80kg"), asks for fitness statistics, or forgets to close an active exercise session.

Metadata

Slug fitness-skill

Version 1.2.2

License MIT-0

All-time Installs 1

Active Installs 1

Total Versions 4

Frequently Asked Questions

What is Fitness & Workout Manager?

运动健康管家 Pro — 专为 Open Claw + 飞书生态打造的 AI 健身管家。双模式打卡（事后总结 / 边练边记）、3 层 Session 防丢失、飞书云文档同步、极致 Token 优化。Route to this skill when the user mentions workouts, logs... It is an AI Agent Skill for Claude Code / OpenClaw, with 380 downloads so far.

How do I install Fitness & Workout Manager?

Run "/install fitness-skill" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Fitness & Workout Manager free?

Yes, Fitness & Workout Manager is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Fitness & Workout Manager support?

Fitness & Workout Manager is cross-platform and runs anywhere OpenClaw / Claude Code is available (linux, darwin).

Who created Fitness & Workout Manager?

It is built and maintained by PH13917403910 (@ph13917403910); the current version is v1.2.2.

More Skills

Fitness & Workout Manager