mrkeyoor.comAsk

spideymooon/McOracletested 15 Nov 2026commit 63875f5

McOracle review

Use McOracle to study a Chinese-language fortune skill that separates random draws from restaurant data.

  • README install 0.7 s
  • README snippet failed
14 sto install
  • 19.8 MB on disk
  • 0 dependencies
AI Toolsevaluationupdated 15 Nov 2026
+0stars / 7d

McOracle is a developer-created fortune-telling skill that combines random Chinese fortune draws, McDonald's China MCP menu queries and programmed card composition. The supplied README and fortune-library documentation are Chinese.

Verdict

Use McOracle to study a Chinese-language fortune skill that separates random draws from restaurant data. Our sandbox installed it in 14 seconds and built it in 5 seconds, but pytest collected no tests. Skip it if you need a demonstrated complete payment and order-status workflow or a worldwide ordering app.

McOracle infographic: draw a fortune and compose a card, plus its uses, our test result, pitfalls and who it suits
Tested 15 November 2026: the README snippet failed. Before that: source install 14 s, build 5 s; uv install took 0.7 s.

We ran it

Lab card: what happened when we ran McOracle
Lab card from our sandbox run
Screenshot of McOracle (github.com/spideymooon/McOracle)
github.com/spideymooon/McOracle
Install✓ · 14s35 packages · 38 MB
Build✓ · 5s
Tests✗ · 6s0 passed · 0 failed of 0 (pytest)
Known vulns0(pip-audit)
Repo30 files~1,362 lines of source · 14.1 MB · 0 CI workflows · tests dir

Answers from our run

Does McOracle build from source?

Dependencies installed in 14 seconds (35 packages), and the build succeeded in 5 seconds. We cloned commit 63875f5 into a clean Debian container with 3 CPUs and no project-specific setup.

Do McOracle's tests pass?

No. The project's own test command finished in 6 seconds and reported 0 passed, 0 failed (pytest).

Does McOracle have known vulnerabilities in its dependencies?

pip-audit found none in the dependency tree at the time of our run.

Who should not use McOracle?

Readers seeking an ordinary worldwide McDonald's ordering app.

What are the alternatives to McOracle?

McDonald's China MCP service, Pillow. Use McOracle to study a Chinese-language fortune skill that separates random draws from restaurant data.

Setup4/5In our 2026-11-15 run it installed as the README says in 0.7 s; the README gave no task we could run as written.Editor’s note: Installation and build succeeded, but the recorded procedures did not complete the connected first-use workflow.
Docs4/5The README has an install section (“环境要求与依赖安装”) and a usage section (“使用示例”); we did not run its quickstart ourselves, and the README is mostly in Chinese, not English.Editor’s note: The supplied Chinese README explains setup and scripts; complete execution of its documented paths is absent from our records.
Community3/5Last push on 2026-10-25 and 609 stars.Editor’s note: The supplied social descriptions and activity snapshots provide no independent customer outcomes or support-resolution history.
Maturity1/5No tagged GitHub release, repository created 2026-10-12, its test suite did not pass in our sandbox and no licence file GitHub recognises.Editor’s note: Our pytest run collected zero tests; the author's separate checks and pending-payment report do not complete the transaction assessment.

How we score

Who it’s for

Developers studying how random entertainment can stay separate from live restaurant data.
Chinese-reading WorkBuddy users interested in this fortune-card experience.
Python developers exploring standalone draw, card and order-state scripts.

Who it’s NOT for

Readers seeking an ordinary worldwide McDonald's ordering app.
Buyers who require a demonstrated complete payment and order-status workflow.
Teams that need established redistribution rights before adopting a project.

Setup reality

Our sandbox run at commit 63875f5 on 2026-11-15 installed successfully in 14 seconds, recorded 35 dependencies and 38 MB on disk, and built successfully in 5 seconds. Tests exited 5 after 6 seconds: pytest recorded 0 passed and 0 failed out of 0, with 'no tests ran in 0.01s'. pip-audit reported 0 known vulnerabilities.

The separate quickstart-lib run at commit 63875f571f23 cloned in 2 seconds, installed Pillow 13.0.0 in 0.7 seconds, recorded 19.8 MB of site-packages and installed 0 packages besides Pillow. Its snippet step failed because the recipe found no Python code block. The supplied README has shell commands invoking Python scripts. Teardown returned ok.

The README requires Python 3.9+ and Chinese fonts on Mac or Linux. Live restaurant queries require an MCP token and an enabled connector; standalone draws have a separate local path.

McOracle installed in 14 seconds and built in 5 seconds in our sandbox, but pytest collected no tests. Use it to study a Chinese-language fortune skill that keeps random entertainment separate from restaurant data. Skip it if you need a demonstrated complete payment and order-status workflow or an ordinary worldwide McDonald's ordering app. The useful design question is how a playful answer becomes a meal suggestion without turning an invented price into a commercial fact.

McOracle assigns those jobs to different components. A Python script draws a fortune from an original library, the restaurant connection supplies menu and price information, and Pillow helps compose the card. The README describes an additional confirmation step before order creation. That division gives developers a concrete system to examine: a stable draw, a recommendation grounded in restaurant responses, a rendered artifact and a separately authorized transaction. McOracle README

What people use McOracle for

The project author reports restaurant queries, 1200×1600 PNG cards and an order returned in pending-payment status. Those examples cover three different reasons to explore McOracle: learning the restaurant connection, producing this particular visual experience or examining order orchestration. The documentation also reports running local checks for the oracle, card renderer and order-state logic. Keeping the author as the subject makes the origin of these results clear without interrupting every example with the same disclaimer.

Who used it What they did Reported outcome
The project author Queried McDonald's China stores, menus, meal details and prices Reports restaurant queries including discounts and collection methods. Author documentation
The project author Composed fortune cards with the programmed template and official product images Reports actual 1200×1600 PNG output. Author documentation
The project author Called the real order-creation interface Reports an order returned in pending-payment status. Author documentation
The project author Ran oracle, card and order-safety checks Reports 24 / 24 passing checks; oracle and order-safety checks use mocks, while card checks produce local PNGs. McOracle test documentation

The README describes a library of 104 Chinese fortunes and a draw locked to the current round. The explanation, card and recommendation refer to that same result until the user asks for another draw. For an assistant developer, this is a useful state-management example: formatting another response should not silently change the underlying fortune. The documented commands distinguish asking a question, inspecting the current draw and requesting a redraw. Oracle architecture and commands

The recommendation rules map fortune tags to food directions, then combine those directions with the available menu and budget. A user who asks only for a fortune stops before the restaurant lookup and location request. That makes the draw script a practical first entry point. Someone interested in restaurant orchestration can instead follow how the recommendation acquires a real item, store and price before those details reach the card or order summary. McOracle project flow

What happened when we ran it

Our run installed McOracle in 14 seconds and built it in 5 seconds at commit 63875f5 on 2026-11-15. Pytest exited 5 after 6 seconds and collected zero tests. The supplied log says 'no tests ran in 0.01s'. These are repository-harness results, recorded separately from the quickstart-lib procedure below. Reading the two procedures side by side preserves both the successful setup steps and the failed attempt to find a runnable snippet.

Harness step or signal Our recorded result
Install Succeeded in 14 seconds; 35 dependencies recorded; 38 MB on disk
Build Succeeded in 5 seconds
Tests Failed with exit 5 in 6 seconds; pytest recorded 0 passed and 0 failed out of 0
Test log no tests ran in 0.01s
Dependency audit pip-audit reported 0 known vulnerabilities
Repository scan 0 CI workflow files, no Dockerfile, a tests directory present

The harness recorded 3 CPUs, 8 GB RAM and the configured image label ghcr.io/astral-sh/uv:python3.12-bookworm. Its checkout contained 30 files, approximately 1,362 source lines and 14.1 MB of files. The recorded total duration was 32 seconds. These figures describe the repository procedure and its configured resources. The test outcome is specifically empty collection: the supplied tail contains neither a failed assertion nor an exception identifying a package or environment cause.

The separate quickstart-lib run used commit 63875f571f23 on 2026-11-15 in a fresh unprivileged container configured with 2 CPUs, 4 GB RAM, no published ports and no secrets. It used the same recorded image label and installed Pillow from PyPI. The following rows retain each supplied step, including the failed snippet and final cleanup, so the shorter install result cannot be confused with the harness's 35-dependency installation.

Quickstart step or finding Recorded outcome
Clone quickstart-lib recipe completed in 2 seconds.
Install uv pip install Pillow into a fresh venv installed Pillow 13.0.0 in 0.7 seconds; the README command was pip install Pillow.
Installed size 19.8 MB on disk in the venv site-packages.
Dependency count 0 packages besides Pillow installed.
Snippet FAILED; the README has no python code block to run.
Finding no-snippet: the README has no python code block to run.
Teardown ok.

The supplied README contains shell-tagged examples that invoke Python scripts, including the draw command and a mock order-controller demonstration. That resolves the apparent contradiction at the documentation level: a recipe looking for a Python code block can stop while a human reader still has shell commands to follow. The failure remains part of our recorded result. The README commands are the next reproducible starting point, rather than a reason to rename that failed step as a successful demonstration. Standalone script examples

Where people get stuck: registration, live data and price units

McOracle's documentation ties WorkBuddy registration metadata to the installed WorkBuddy version and provides independently runnable scripts. That gives a developer two routes into the project: load the skill through WorkBuddy or start with the local scripts. For restaurant queries, the README adds a token, connector configuration and an enabled connection. For card rendering on Mac or Linux, it asks for an available Chinese font. These are separate setup tasks with different symptoms. McOracle setup requirements

Situation Documented response What to check first
WorkBuddy does not load the skill Supply metadata required by the installed version; scripts can run independently Local WorkBuddy registration requirements
MCP is not configured Oracle draws and product-free cards remain available; recommendations degrade Whether the requested experience needs live restaurant data
A price lacks live confirmation Display a price-unchecked label The restaurant response and its timestamp
Order details change after confirmation Discard the old confirmation and request another Item, quantity, store, price and collection method

The documentation says restaurant prices arrive in fen and must be divided by 100 for display. It also distinguishes pickup and drive-through parameters from delivery parameters. Those details belong in the data flow: keep the selected collection method attached to the price calculation and the order summary. A renderer should receive the price's status as well as its amount, so a template example cannot accidentally become a claim about the customer's current meal. Integration and price-unit notes

A sensible reading order follows those component boundaries. Start with the draw script and its saved round state, then inspect card composition and font selection. Add the restaurant connector when the goal becomes live menu recommendations. This sequence comes from the project's documented split between standalone scripts and skill orchestration. It also makes troubleshooting more specific: a script invocation, a font problem and WorkBuddy registration each have their own place to investigate. McOracle architecture and setup

Confirmation is part of the order data

The README says a change to the item, quantity, store, price or collection method invalidates the previous order confirmation. Approval therefore belongs to a particular order summary. The documented conversation moves from preparing an order to showing its details, receiving explicit confirmation, checking the price again and creating the order. For developers adapting the flow, preserving those states matters as much as preserving a product identifier or quantity. Order-safety rules

The project distinguishes order creation from payment and asks the assistant to return a real order number and official payment entry point. Its mock controller is presented as pure logic, while the restaurant connector performs commercial operations. Tokens belong in environment variables or connector configuration under the README's rules. Contributors are also asked to exclude tokens, order numbers and personal information from issues and pull requests. These are documented operating rules for readers to examine when adapting the skill. Order flow and contributor instructions

Chinese text and McDonald's China define the audience

The supplied README and fortune-library documentation are Chinese. The documented library has 104 entries across 8 categories, and the conversational triggers are Chinese phrases such as requests for a fortune or another draw. Card text is composed with Pillow, and the setup instructions ask Mac and Linux users to provide a Chinese font. This makes language and font availability part of the first-use decision for an English-reading developer. McOracle language and rendering documentation

The official McDonald's China MCP guide describes a service for mainland China, excluding Hong Kong, Macao and Taiwan. It provides the restaurant interface McOracle connects to and links an English guide. McOracle itself is described as a developer contest entry rather than an official McDonald's product. Keep those identities clear: the official restaurant service supplies the connection, while the developer's skill supplies the fortune experience and its orchestration. Official MCP service guide, McOracle project declaration

Stars show repository attention

Trendshift records 481 additional stars, 24 additional forks and zero merged pull requests over its stated October 12 to November 15, 2026 period. These figures describe attention and repository activity. GitHub metadata separately records 609 stars, zero open issues and PRs, and a last push on October 25, 2026 when observed on November 15. The two snapshots answer different questions: total attention at a point in time and changes during a stated interval. Trendshift activity record, Supplied GitHub metadata

A post by vintcessun on X describes Python library draws, MCP menu data, programmed cards and explicit order confirmation. That is a useful description of the project's intended division of work. Entity names also matter here: MiCAOracle, MacroOracle and a Musicboard account named mcoracle concern different entities. The review's subject is spideymooon/McOracle, the Chinese fortune skill, and its component choices follow that specific purpose. vintcessun's McOracle post, McOracle repository

McOracle vs the official MCP connection and Pillow

The official McDonald's China MCP service supplies restaurant tools, while Pillow supplies Python image processing. McOracle combines those kinds of building blocks with its own fortune library, recommendation rules and conversation instructions. The choice is therefore about which application behavior you want to adopt and which you want to write. A developer interested in restaurant access can begin with the official interface; a developer interested in a different card design can begin with the image library.

Choice What its documentation supplies Pick it when
McOracle A fortune library, recommendation rules, card scripts and order-safety instructions You want this particular Chinese-language entertainment experience
McDonald's China MCP service An official MCP data interface for mainland China You want to write your own restaurant assistant around the official connection
Pillow Python image processing capabilities You want to build your own card renderer and surrounding behavior

The official service guide documents a remotely hosted MCP endpoint, Streamable HTTP and token-based authorization. Starting there means designing the surrounding conversation and confirmation behavior yourself. Pillow's documentation describes image-processing capabilities for a Python interpreter; using it directly means creating your own layout and application logic. These are concrete differences in responsibility, which make the table useful even for a developer who intends to borrow only one part of McOracle's design. Official restaurant interface, Pillow overview

What the other reviews get wrong

ToolPotion lists AI-generated responses and categorizes McOracle as cloud deployment on the web platform. McOracle's README instead describes a randomly selected fortune from an original library, a skill copied into local WorkBuddy directories and independently runnable Python scripts. An assistant's interpretation of a draw is distinct from selecting that draw. Likewise, a connection to a remote restaurant service is distinct from operating the surrounding skill as a hosted website. ToolPotion listing, McOracle architecture and installation

Those distinctions change the setup a reader should expect. The local skill directory, Python commands, Chinese fonts and connector configuration are part of the documented experience. A repeated fortune can be a valid library result, while a repeated example price has a different meaning because restaurant data changes. Reading the source documentation gives a more specific basis for choosing the project than a generic listing's deployment and feature labels. McOracle README

FAQ

McOracle's README requires explicit confirmation before order creation and leaves payment to the user on the official page. Its FAQ also addresses price differences, use without MCP, image-generation keys, entertainment boundaries and card typography. The six questions below come from that supplied documentation. Their answers describe the project's stated behavior and give first-time readers a direct route to the relevant setup or operating rule. McOracle FAQ

Will McOracle place an order or charge me automatically?

The README requires explicit order confirmation and says McOracle does not pay automatically. Asking to look at food, receive a recommendation or prepare an order does not count as confirmation. After order creation, the user pays on the official page. The order-safety rules also require a new summary and confirmation when relevant details change, keeping consent attached to the selected meal and collection method. Order-confirmation FAQ

Why might McOracle's prices differ from the McDonald's app?

The FAQ directs readers to the selected store's live query and says the card and conversation display a confirmation time. Its answer treats that response as the price reference rather than providing a diagnosis for every possible app mismatch. Under the documented rule, a price lacking live confirmation receives the price-unchecked label. The README's sample meal and amount illustrate a conversation; the selected store's response supplies the customer's price. Price FAQ

Can I use McOracle without configuring MCP?

Yes: the documentation says oracle draws and product-free fortune cards work without MCP. Product recommendations degrade and use the price-unchecked label. This local path suits someone interested in the fortune experience or card composition. The restaurant connection becomes the next setup step when the goal includes live menus and prices, rather than a requirement for reading or drawing from the fortune library. Use-without-MCP FAQ

Does McOracle need an image-generation API key?

No image-generation API key is required for standard card output, according to the README. The normal card uses programmed layout with product imagery or the documented fallback artwork. An optional background-preview mode is described separately for visual experiments. The distinction helps readers choose the intended workflow: the ordinary card path starts with the supplied template and rendering script, without selecting an image-generation provider. Card-generation FAQ

Does McOracle predict the future?

No: the README describes the fortunes as entertainment and inspiration rather than a basis for real-world decisions. It says major subjects such as resignation, investment, health and law receive a reminder to exercise independent judgment. The random draw is the premise of the experience, and the library gives that draw its wording. The meal suggestion is a separate recommendation step grounded in restaurant data. Fortune-purpose FAQ

Will Chinese text become garbled or overflow on McOracle's cards?

The README describes automatic wrapping and font-size reduction through Pillow, plus a font containing the yuan glyph for prices. It asks Mac and Linux users to ensure a Chinese font is present and describes candidate-font selection by the script. Those instructions provide specific rendering checks for a new machine: start with available fonts, then inspect the resulting text and price display in the generated artifact. Typography FAQ

What we have not tested yet

Our supplied runs contain no completed restaurant query, generated card, paid order or order-status retrieval. The harness recorded repository setup and empty pytest collection; the separate quickstart stopped at its snippet step. The cause of empty collection remains unresolved, and the configured image label does not establish actual runtime or operating-system versions. The author's reported 24 / 24 checks are separate from our pytest result. No corrected or additional example was executed for this revision.

The author's documentation reports an order returned in pending-payment status, with payment and subsequent status retrieval outside the completed demonstration. It also lists delivery, coupon collection and redemption, and the visual appearance of fallback concept illustrations among its unfinished checks. Our evidence contains no comparative performance measurements, migration outcomes or independent customer results. Stars, forks and descriptive social posts therefore supply attention and presentation evidence rather than transaction reliability or user satisfaction. Author capability statement

The supplied README and fortune-library documentation are Chinese; English setup documentation was not verified. The supplied metadata does not establish a latest release. GitHub's license field is empty, and the README's learning-and-demonstration statement does not establish redistribution rights. Our records also leave token handling, privacy behavior, confirmation bypasses, concurrent users, sustained volume and service rate-limit behavior untested. The clean dependency audit addresses known package advisories in the recorded scan, rather than those application behaviors.

The supplied pack contains no relevant independent GitHub issue investigation or McOracle video walkthrough, and it provides no basis for judging support-resolution speed. Rendered review assets and other site-level presentation elements are also outside this text correction. The next useful evidence would be a retained live menu response, a card generated on the intended machine and a separately authorized transaction followed through payment and status retrieval. Developers can use the existing material to study the architecture while keeping those remaining questions explicit.

Alternatives

ProjectWhat it isPick it when
McDonald's China MCP serviceThe official restaurant interface used by McOracle.Choose this as a building block when you want to write your own restaurant assistant without the fortune experience.
PillowA Python image-processing library used in McOracle's card composition.Choose this as a building block when you want to design your own card renderer and surrounding application.

Sources

  1. McOracle README and source
  2. Author documentation mirrored by LobeHub
  3. Trendshift activity snapshot
  4. ToolPotion listing
  5. X discussion
  6. Official McDonald's China MCP guide
  7. Pillow documentation
  8. Supplied GitHub repository metadata

More ai tools reviews