McPerks Planner skill
A WorkBuddy skill over the McDonald's MCP that ranks coupons and points by urgency and checks prices against the server.
- 121 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The skill reads account data through an MCP token; orders and redemption run only after confirmation.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/mcperks-plannerDetects 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 skill reads account data through an MCP token; orders and redemption run only after confirmation.
This is third-party code. Review the repository files before installing.
What it does
The skill takes data from the official McDonald's MCP and turns it into a clear plan. The inventory module gathers coupons, claimable coupons, points and campaigns and sorts them into usable, expiring and no-expiry. The planning module ranks offers by urgency and a conservative savings estimate within a one-day or multi-day budget. The pricing module takes the final price from the `calculate-price` response, converting cents to yuan, and separates the estimate from the actual total. The campaigns module links events to coupons and dishes and marks future campaigns as awaiting confirmation. The algorithm core has no network dependency, so results are reproducible. The commands `inventory`, `plan`, `reprice`, `campaigns`, `compare` and `demo` also run without WorkBuddy. By default the skill only reads and plans; ordering, claiming a coupon and redeeming ask the human first.
Who it is for. McDonald's users in China and developers who want an example of a skill with deterministic calculations on top of MCP.
Good fit when
- Coupons and points are nearly expired and need ordering by urgency.
- Within a weekly budget, offers with the best conservative savings are picked.
- A recommended price differs from the checkout price and an exact calculation is needed.
- The skill demo is run without a token using `demo`.
Not a fit when
- A service outside China is needed.
- An agent would place orders without human confirmation.
- An official McDonald's product is expected: this is an independent contest entry.
Example request
Show coupons and points by urgency, build a three-day plan with a 100 yuan budget and check the price through `calculate-price`.Limitations
The project is a programmer contest entry, not a McDonald's product, and its output is for reference. It targets WorkBuddy and the Chinese MCP, with Chinese documentation. Real data needs an MCP token kept only in connector settings. Some campaign dates are unknown and such items are marked as pending. The direct MCP client in `mcp_adapter.py` is optional. Running tests needs `pytest`; everything else uses the Python 3.12 or newer standard library. The last commit dates from October 2026.
How to disable. Disable the MCP connector in WorkBuddy and delete the skill folder. The token lives in connector settings and is revoked in the MCP console.
FAQ
Can I try it without a token?
Yes, `python -m mcperks.cli demo` shows a sample run without a token or network calls.
How does it tell an estimate from the real price?
The real total comes from the `calculate-price` response and the estimate is marked separately, which exposes the gap between recommendation and checkout.
What can the command line do?
The subcommands `inventory`, `plan`, `reprice`, `campaigns`, `compare` and `demo` read JSON input and print results.
Does the skill place orders on its own?
No. By default it only queries and plans, and orders, coupons and redemption need human confirmation.
Is it an official product?
No, it is an independent developer's contest entry. Current prices and availability come from official McDonald's responses.
Discussion
No questions or reviews yet.
Related
SkillsA WorkBuddy skill for the McDonald's MCP: it computes coupons and points, assembles orders and guards draws with a ledger.
A WorkBuddy skill reads McDonald's coupons and points through MCP and proposes up to two price-verified orders within a budget.
A WorkBuddy skill that totals the calories in a month of McDonald's orders and assembles a healthier next lunch.
This skill prices a real McDonald's group order by city, headcount and budget through the vendor's own official MCP.
A skill for ordering from McDonald's China via its official MCP: coupon-aware prices, nutrition, order status and a quiet summary.
A WorkBuddy skill that queries stores and coupons through McDonald's China's official MCP and finds the cheapest meal combination.
README in short
The architecture is described in layers: conversation via `SKILL.md`, MCP calls, response parsing and deterministic calculation. The author avoids databases, Docker and Redis.