Cook Like Hoc

Search and display recipes from the open-source Gar-b-age/CookLikeHOC repository. Use when the user asks to "cook like hoc", "像老乡鸡那样做饭", "老乡鸡菜谱", "做菜", or wa...

Security Scan

The skill does what it says: a small Python helper that looks up recipes from the named GitHub repo and returns markdown; nothing in the code or instructions appears disproportionate or unrelated to that purpose.

Purpose & Capability

Name/description match implementation: the script fetches the repo tree from GitHub and recipe markdown via the jsDelivr CDN and returns recipe content. The requested capabilities (network fetches of the repository) are appropriate for the stated purpose and there are no unrelated credentials or binaries required.

Instruction Scope

SKILL.md instructs the agent to run the included Python script from the skill folder. The script only reads/writes a cache file inside the skill directory, queries the GitHub tree API, and fetches raw markdown via jsDelivr. It does not attempt to read unrelated system files or send data to unexpected endpoints.

Install Mechanism

No install spec; this is instruction/code-only. There are no remote installers or downloads performed by an installer step. The only network activity is the script fetching repository data at runtime (GitHub API and jsDelivr), which is expected.

Credentials

The script will optionally read a GITHUB_TOKEN environment variable (os.environ.get('GITHUB_TOKEN')) and add it to the Authorization header to increase API rate limits. The skill metadata does not declare any required env vars. This is plausible for the purpose but worth noting: if a token is present it will be sent to api.github.com (normal GitHub usage).

Persistence & Privilege

always:false and no system-wide config changes. The script writes/updates a cache file (recipe_tree_cache.json) within the skill's scripts directory only; it does not modify other skills or global agent settings.

Assessment

This skill is coherent for fetching recipes from the referenced GitHub repo. Before installing, be aware that: (1) the script makes outbound HTTP(S) requests to api.github.com and to jsDelivr (fastly.jsdelivr.net) to list and fetch files; (2) if you have a GITHUB_TOKEN environment variable set, the script will read it and use it in an Authorization header to GitHub (this is optional and used only to raise API rate limits) — remove or limit that token if you don't want it used; (3) the script writes a local cache file (recipe_tree_cache.json) inside the skill directory and may update it. If those behaviors are acceptable, the skill appears safe and consistent with its description. If you want extra caution, inspect the included files yourself or run the script in a sandboxed environment (no token, restricted network) first.

CookLikeHOC Skill

This skill allows you to retrieve recipe instructions from the CookLikeHOC repository. It contains recipes from the "老乡鸡菜品溯源报告".

Prerequisites

• Python 3 installed
• Built-in urllib and json modules (no external libraries required)

Workflows

1. Searching for a Recipe

When the user asks for a recipe or wants to know how to cook a dish from the CookLikeHOC database, run the search_recipe.py script.

Command:

python ~/.agents/skills/cook-like-hoc/scripts/search_recipe.py "<Recipe Name>"

Example:

python ~/.agents/skills/cook-like-hoc/scripts/search_recipe.py "肥肠鸡"

How to handle the output

• The Python script performs fuzzy matching. It will return the markdown recipe for the exact or closest match.
• Output the recipe to the user, formatting it cleanly as markdown.
• If it says Could not find any recipe matching, inform the user that the recipe isn't in the database and perhaps suggest one from the provided examples.

Supported categories & recipes

The repository has categories like 主食 (Staple Food), 炒菜 (Stir-fried), 汤 (Soup), 炖菜 (Stew), 炸品 (Fried), 烤类 (Roasted), 烫菜 (Blanched), 煮锅 (Boiled Pots), 砂锅菜 (Casserole), 蒸菜 (Steamed), 配料 (Ingredients), and 饮品 (Drinks). The script searches across all markdown files in these categories automatically.