Official Verified

recipe-scout

Find and normalize Chinese recipes (中餐菜谱) from structured sources first, then export clean recipe notes to Obsidian markdown. Prefer stable recipe sites over social posts; use 小红书 only for inspiration/fallback. Use when user asks for: Chinese recipes / 中餐菜谱, "how to cook X" / 做法, meal ideas, home-style dishes, weeknight cooking, recipe collection / curation, or recipe note creation in Obsidian.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/cylqqqcyl/recipe-scout

Download Source Code (.zip)

recipe-scout

Source-policy + normalization layer for Chinese recipes. Not a single-site scraper.

Source Ranking (strict priority)

Tier A — Structured recipe pages (primary)

下厨房 (xiachufang.com) — largest Chinese UGC recipe platform, step-by-step photos, strong search
美食天下 (meishichina.com) — broad category coverage, reliable ingredient lists
豆果美食 (douguo.com) — large recipe app ecosystem
HowToCook (github.com/Anduin2017/HowToCook) — programmer-authored, precise gram-level measurements, no fluff; raw URLs: https://raw.githubusercontent.com/Anduin2017/HowToCook/master/dishes/<category>/<dish>/<dish>.md; categories: meat_dish, aquatic, vegetable_dish, staple, soup, dessert, condiment
Reputable cooking blogs with full recipe text

Tier B — Video + transcript (secondary)

YouTube / Bilibili cooking videos with captions or clear on-screen ingredients

Tier C — Social posts (fallback / inspiration only)

小红书, short-post platforms, comment threads
Mark confidence low unless validated by another source
Do NOT treat as authoritative recipe specs

Anti-pattern: Do not make 小红书 the primary recipe backend. UI changes often, ingredient/timing details are frequently missing, extraction is brittle.

Retrieval Strategy

Parse dish intent + constraints (cuisine, diet, equipment, time, restrictions)
Search Tier A first with Chinese queries (see references/query-examples.md)
Collect 3–7 candidates
Deduplicate by core technique + ingredient profile
Rank by: completeness → clarity → home-cooking fit → ingredient accessibility
Use Tier B/C only to fill gaps or add variants

Normalization Rules

Language: Chinese default (English if user requests)
Units: metric preferred (g / ml / tbsp / tsp); preserve source units if conversion uncertain
Times: separate prep / cook / total; mark missing estimates as estimated: true
Seasoning: keep exact source values; record 适量 as-is with a practical note
Heat: normalize to 小火 / 中小火 / 中火 / 中大火 / 大火
Safety: include basic food safety notes (chicken/pork doneness, etc.) when relevant

Anti-Hallucination

Never fabricate ingredient quantities, temperatures, or timings
Missing fields → unknown / 未注明
Clearly distinguish: source facts vs inferred estimates vs your recommendations

Output Modes

Mode	When	Output
Quick answer	Casual ask	Top 3 options, recommended version, concise steps + shopping list
Structured pack	User wants detail	Normalized recipe using schema
Obsidian export	Collection/save	One `.md` file per recipe → `/home/node/vault/Recipes/Chinese/`; fallback: `{workspace}/exports/recipes/`

Schema: See references/schema.md Obsidian template: See references/obsidian-template.md

Execution Playbook

Read Full Documentation on GitHub

Metadata

Author@cylqqqcyl

Stars3409

Updated2026-03-25

View Author Profile

AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill

Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-cylqqqcyl-recipe-scout": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Safety NoteClawKit audits metadata but not runtime behavior. Use with caution.

Related Skills

body

Physical performance — workouts, fitness tracking, nutrition, meal logging, macros, recipes. Use when the user mentions exercise, lifting, running, training, food, eating, macros, protein, meal planning, or anything related to physical health and performance. Routes to sub-skills: coach (movement/training) or chef (nutrition/food).

cylqqqcyl 3409

lean

Run QuantConnect LEAN backtests and manage US equity algorithm development. Use when asked to backtest a trading strategy, run a LEAN algorithm, analyze backtest results, download market data, or deploy to Interactive Brokers TWS. Covers algorithm creation, data management, config editing, and result analysis.

cylqqqcyl 3409