ClawBee — Family Meal Planner Skills for OpenClaw

Photo your fridge → get a 3-day meal plan → receive the shopping list on Telegram.

---

Architecture

flowchart TD
    User(["👤 User"])

    TG["📱 Telegram Bot\n─────────────\nSend fridge photos\nReceive shopping lists"]
    DC["🎮 Discord\n─────────────\nView meal plan posts\nBot-to-bot updates"]

    subgraph RAIL ["🚂 Railway Cloud"]
        subgraph OC ["🦞 OpenClaw Gateway"]
            direction TB
            ORCH["🚀 Orchestrator\n/plan weekly · /plan post"]

            subgraph SKILLS ["Agent Skills  ·  npx skills add mary4data/ClawBee"]
                direction LR
                FS["📸 Fridge Scanner\n/scan · /scan demo"]
                FT["🧊 Fridge Tracker\n/fridge list · add"]
                MP["📅 Meal Planner\n/meals plan"]
                PH["💰 Price Hunter\n/prices search"]
                SA["🛒 Shopping Agent\n/shopping send"]
            end

            ORCH --> FS & FT & MP & PH & SA
        end

        DB[("🗄️ SQLite\npantry.db")]
    end

    FAI["🤖 Featherless AI\nLLM Backend"]
    WEB["🌐 Web Search\nRewe · Lidl · Aldi Berlin"]

    User -->|"fridge photo / command"| TG
    User -->|"views posts"| DC
    TG -->|"inbound message + media"| OC
    OC -->|"shopping list"| TG
    OC -->|"meal plan post"| DC

    FS & FT & MP & PH & SA <-->|"read / write"| DB
    FS & MP -->|"LLM inference"| FAI
    PH -->|"price search"| WEB
    SA -->|"send list"| TG
    SA -->|"notify"| DC

    classDef channel  fill:#5865F2,color:#fff,stroke:#4752C4
    classDef skill    fill:#1e293b,color:#94a3b8,stroke:#334155
    classDef external fill:#0f172a,color:#64748b,stroke:#1e293b
    classDef db       fill:#713f12,color:#fde68a,stroke:#92400e
    classDef user     fill:#166534,color:#bbf7d0,stroke:#15803d
    classDef orch     fill:#7c3aed,color:#ede9fe,stroke:#6d28d9

    class TG,DC channel
    class FS,FT,MP,PH,SA skill
    class FAI,WEB external
    class DB db
    class User user
    class ORCH orch

Loading

---

Setup on Railway (5 min)

Step 1 — Create a Telegram Bot

1. Open Telegram → search @BotFather
2. Send /newbot and follow the prompts
3. Copy the token: 123456789:ABCDEF...
4. Get your Chat ID: message @userinfobot → copy the Id number

Step 2 — Set Railway Environment Variables

In Railway → your OpenClaw service → Variables, add:

Variable	Value
TELEGRAM_BOT_TOKEN	123456789:ABCDEF...
TELEGRAM_CHAT_ID	123456789
FEATHERLESS_API_KEY	your Featherless key
DISCORD_BOT_TOKEN	(optional)
DISCORD_CHANNEL_ID	(optional)

Step 3 — Run Auto-Setup

Open the Railway shell (your service → Shell tab) and run:

bash <(curl -s https://raw.githubusercontent.com/mary4data/ClawBee/main/deploy/setup.sh)

This single command:

• Writes openclaw.json with Telegram + Featherless AI config
• Clones ClawBee skills into /data/workspace/clawbee/skills
• Creates pantry.db with all tables
• Prints test commands when done

Step 4 — Restart OpenClaw

In Railway → your service → Restart (or redeploy).

Step 5 — Test

Send to your Telegram bot:

/plan help

---

Install Skills Only

If OpenClaw is already configured, just install the skills:

npx skills add mary4data/ClawBee

Or individual skills:

npx skills add mary4data/ClawBee@orchestrator
npx skills add mary4data/ClawBee@fridge-scanner

---

Commands

Command	Description
/plan weekly [budget]	Full pipeline — fridge check, prices, meal plan, Telegram
/plan post	Post plan to Discord
/scan + photo	Scan fridge photo → 3-day plan
/scan demo	Demo scan (no photo needed)
/scan shop	Send scan's shopping list to Telegram
/fridge list	Show pantry contents
/fridge add <item> [qty]	Add item
/meals plan [budget]	7-day dinner plan
/prices search <item>	Find cheapest price in Berlin
/shopping send	Send shopping list to Telegram
/shopping optimize [€]	Check against budget

---

Manual Config

See deploy/openclaw.json.example for the full config file. Place at /data/.openclaw/openclaw.json on your Railway instance.

---

Stack

Component	Technology
Gateway	OpenClaw on Railway
LLM	Featherless AI
Messaging	Telegram Bot API + Discord
Storage	SQLite (pantry.db)
Skills	skills.sh/mary4data/ClawBee

---

License

Apache 2.0 — see LICENSE.txt