mcd-order-calories skill
Calories for McDonald's China orders from MCP: combos split into items, names matched to the table, and gaps kept visible.
- 100 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The token grants access to account order history.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/mcd-order-caloriesDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The token grants access to account order history.
This is third-party code. Review the repository files before installing.
What it does
The skill reads order history through the McDonald's China MCP server and totals calories per order. Combos are expanded into real items, and names are matched against the official nutrition table in several stages. The `mcd-calories` command prints each order's energy, the share of matched items and the source of every number. Items missing from the table are marked as uncounted instead of hiding inside an understated total. It is read-only: only two queries are used, an order list and a nutrition list.
Who it is for. McDonald's China app users who track nutrition and want calorie counts for real orders.
Good fit when
- You want weekly or monthly calories from past orders.
- An agent is asked to split a combo into items and count each one.
- You check which products are missing from the nutrition table.
- You compare orders by day and by restaurant.
Not a fit when
- You have no access to the McDonald's China MCP and its token.
- You need to place an order or claim a coupon: the tool writes nothing.
- You need a medical diet calculation.
Example request
Total the calories of my McDonald's orders for September and show which items were not found in the nutrition table.Limitations
A McDonald's China MCP server token is required in `MCD_MCP_TOKEN`; it is issued by the M-China/mcd-mcp-server project. Data and names are in Chinese, and so is the command output. Some items are estimated from composition with a confidence label, and unmatched ones stay uncounted. The token grants access to your account order history, so keep it out of repositories and do not share it. The project was built for a contest and is still small, so its future depends on the author.
How to disable. Remove the skill from the agent's skills directory, run `python3 -m pip uninstall` for the package and drop `MCD_MCP_TOKEN` from the environment.
FAQ
What does the command do?
`mcd-calories` prints calories per order and a coverage summary. The `--detail` flag shows order contents and where each number comes from.
How do I set the token?
Export the `MCD_MCP_TOKEN` variable. The token is requested from the M-China/mcd-mcp-server project.
Does it install dependencies?
No. Python 3.9 or newer is needed and the project requires no third-party packages.
Does it change orders?
No. Only `order-list` and `list-nutrition-foods` are called, which is read access.
What if an item is not in the table?
It stays on the uncounted list, and the total reports coverage instead of an understated sum.
Discussion
No questions or reviews yet.
Related
SkillsA skill and CLI for the McDonald's China menu: three meal options for a budget, with nutrition and a source for every figure
CLI and skill for McDonald's China orders: pick a meal by calories, price it with coupons and hold the order until confirmed.
A WorkBuddy skill that totals the calories in a month of McDonald's orders and assembles a healthier next lunch.
McButler (mcd-self-butler)
mcd-self-butler
McButler works out calories and the full cost of an order, including travel and queue time, from the McDonald's China MCP
Picks a McDonald's China order for budget, calories, party size and restrictions, claiming coupons first and verifying prices.
A skill for ordering from McDonald's China via its official MCP: coupon-aware prices, nutrition, order status and a quiet summary.
README in short
The README describes five matching levels and promises not to pass a lower bound off as the full total. Tests and data ship in the repository.