← 返回 Skills 市场
1305
总下载
2
收藏
4
当前安装
2
版本数
在 OpenClaw 中安装
/install food
功能描述
Your intelligent food system. Absorbs, analyzes, and organizes everything you eat.
使用说明 (SKILL.md)
Intelligent Food Absorption
This skill absorbs ANY food input, auto-classifies it, and organizes for insights.
Rules:
- Auto-detect input type: meal photo, nutrition label, recipe, menu, text
- Extract and structure: items, portions, context, nutrition when visible
- Tag everything: #meal, #recipe, #product, #restaurant, #inventory
- Offer analysis: "Want nutrition estimate?" — don't force it
- Build personal database: scanned labels, frequent meals, saved recipes
- Provide insights: patterns, variety, timing, correlations
- Remember restrictions permanently, flag conflicts proactively
- For detailed macro tracking → complement with
caloriesskill - Check
processing.mdfor how each input type is handled
Memory Storage
All user data persists in: ~/food/memory.md
Format:
### Preferences
\x3C!-- Their food preferences and restrictions. Format: "item: type" -->
\x3C!-- Examples: nuts: allergy, gluten: intolerance, vegetarian: choice -->
### Products
\x3C!-- Scanned/saved products for quick-log. Format: "product: cal/serving" -->
\x3C!-- Examples: Hacendado yogurt: 120/170g, Oatly oat milk: 45/100ml -->
### Patterns
\x3C!-- Detected eating patterns. Format: "pattern" -->
\x3C!-- Examples: breakfast ~8am, snacks after 10pm, eats out Fridays -->
### Places
\x3C!-- Restaurants and spots. Format: "place: notes" -->
\x3C!-- Examples: Noma: loved fermented plum, Local Thai: go-to takeout -->
### Recipes
\x3C!-- Saved recipes. Format: "dish: key info" -->
\x3C!-- Examples: quick hummus: chickpeas+tahini+lemon 5min, Sunday roast: 2h -->
Empty sections = no data yet. Absorb, classify, organize.
Insights provided: Weekly variety score, meal timing patterns, frequent foods, eating out ratio, nutrition estimates when asked. Not medical advice.
安全使用建议
This skill is coherent with a local food tracker, but before installing you should: (1) Confirm how images and OCR are processed — locally or uploaded to external services — because SKILL.md does not specify network endpoints; (2) Be aware the skill will write a persistent file at ~/food/memory.md containing preferences, allergies, and other personal data; ask whether that file is encrypted, how to delete or export it, and who (which agents/processes) can read it; (3) If you store sensitive allergies or medical info, consider limiting what you save or use a sandboxed account; (4) Test with non-sensitive sample data to verify behavior; (5) If you need offline-only behavior, verify there are no hidden network calls before trusting real photos/labels. If any of these answers are unsatisfactory, avoid enabling the skill or remove the persisted file after testing.
功能分析
Type: OpenClaw Skill
Name: food
Version: 1.0.1
The skill bundle defines a food tracking system that processes user input, extracts food-related information, and stores it in a local markdown file (`~/food/memory.md`). All instructions to the AI agent in `SKILL.md` and `processing.md` are aligned with the stated purpose of absorbing, analyzing, and organizing food data. There is no evidence of malicious intent, unauthorized data access, exfiltration, command execution, or prompt injection attempts designed to subvert the agent's core functions or compromise the system.
能力评估
Purpose & Capability
Name and description align with the instructions: the skill classifies meals, tags items, and builds a persistent personal food database. However, the SKILL.md assumes image processing (OCR, barcode reading, photo item identification) and nutrition estimation capabilities without declaring any required binaries, libraries, or external services — this is an operational gap (not necessarily malicious) that affects how the skill will actually work.
Instruction Scope
Instructions direct the agent to persist all user data (preferences, scanned labels, recipes, patterns, places) to a specific path (~/food/memory.md) and to 'remember restrictions permanently.' There are no instructions about consent prompts, data deletion, encryption, or export controls. Persisting sensitive info (allergies, dietary restrictions, photos/labels) without explicit privacy controls is a privacy risk.
Install Mechanism
No install spec and no code files — the skill is instruction-only, so nothing is downloaded or written by an installer. This minimizes supply-chain risk. The static scanner had no files to analyze beyond the SKILL.md and processing.md.
Credentials
The skill requests no environment variables, credentials, or external config paths. That is proportionate to its described local-tracking purpose.
Persistence & Privilege
The skill does not request elevated platform privileges and isn't set to always:true. It does, however, request persistent local storage in the user's home directory (~/food/memory.md). Persisting personal health-related data locally is consistent with a tracker but increases exposure if the file is unencrypted or accessible by other processes/users.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install food - 安装完成后，直接呼叫该 Skill 的名称或使用
/food触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.1
Preferences now persist across skill updates
v1.0.0
Initial release
元数据
常见问题
Food Tracker 是什么？
Your intelligent food system. Absorbs, analyzes, and organizes everything you eat. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 1305 次。
如何安装 Food Tracker？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install food」即可一键安装，无需额外配置。
Food Tracker 是免费的吗？
是的，Food Tracker 完全免费（开源免费），可自由下载、安装和使用。
Food Tracker 支持哪些平台？
Food Tracker 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Food Tracker？
由 Iván（@ivangdavila）开发并维护，当前版本 v1.0.1。
推荐 Skills