“cpp-programming-guidelines” is a ready-made rule for General, built for Cursor. Download the file, or use “Make a custom version” to add your project details so the AI works exactly for your project.
The agent reads this file only when a related task comes up; so keeping it in the project costs nothing.
Always declare the type of each variable and function (parameters and return value).
Create necessary types and classes.
Use Doxygen style comments to document public classes and methods.
Don't leave blank lines within a function.
Follow the one-definition rule (ODR).
Nomenclature
Use PascalCase for classes and structures.
Use camelCase for variables, functions, and methods.
Use ALL_CAPS for constants and macros.
Use snake_case for file and directory names.
Use UPPERCASE for environment variables.
Avoid magic numbers and define constants.
Start each function with a verb.
Use verbs for boolean variables. Example: isLoading, hasError, canDelete, etc.
Use complete words instead of abbreviations and ensure correct spelling.
Except for standard abbreviations like API, URL, etc.
Except for well-known abbreviations:
i, j, k for loops
err for errors
ctx for contexts
req, res for request/response parameters
Functions
Write short functions with a single purpose. Less than 20 instructions.
Name functions with a verb and something else.
If it returns a boolean, use isX or hasX, canX, etc.
If it doesn't return anything (void), use executeX or saveX, etc.
Avoid nesting blocks by:
Early checks and returns.
Extraction to utility functions.
Use standard library algorithms (std::for_each, std::transform, std::find, etc.) to avoid function nesting.
Use lambda functions for simple operations.
Use named functions for non-simple operations.
Use default parameter values instead of checking for null or nullptr.
Reduce function parameters using structs or classes
Use an object to pass multiple parameters.
Use an object to return multiple results.
Declare necessary types for input arguments and output.
Use a single level of abstraction.
Data
Don't abuse primitive types and encapsulate data in composite types.
Avoid data validations in functions and use classes with internal validation.
Prefer immutability for data.
Use const for data that doesn't change.
Use constexpr for compile-time constants.
Use std::optional for possibly null values.
Classes
Follow SOLID principles.
Prefer composition over inheritance.
Declare interfaces as abstract classes or concepts.
Write small classes with a single purpose.
Less than 200 instructions.
Less than 10 public methods.
Less than 10 properties.
Use the Rule of Five (or Rule of Zero) for resource management.
Make member variables private and provide getters/setters where necessary.
Use const-correctness for member functions.
Exceptions
Use exceptions to handle errors you don't expect.
If you catch an exception, it should be to:
Fix an expected problem.
Add context.
Otherwise, use a global handler.
Use std::optional, std::expected, or error codes for expected failures.
Memory Management
Prefer smart pointers (std::unique_ptr, std::shared_ptr) over raw pointers.
Use RAII (Resource Acquisition Is Initialization) principles.
Avoid memory leaks by proper resource management.
Use std::vector and other standard containers instead of C-style arrays.
Testing
Follow the Arrange-Act-Assert convention for tests.
Name test variables clearly.
Follow the convention: inputX, mockX, actualX, expectedX, etc.
Write unit tests for each public function.
Use test doubles to simulate dependencies.
Except for third-party dependencies that are not expensive to execute.
Write integration tests for each module.
Follow the Given-When-Then convention.
Project Structure
Use modular architecture
Organize code into logical directories:
include/ for header files
src/ for source files
test/ for test files
lib/ for libraries
doc/ for documentation
Use CMake or similar build system.
Separate interface (.h) from implementation (.cpp).
Use namespaces to organize code logically.
Create a core namespace for foundational components.
Create a utils namespace for utility functions.
Standard Library
Use the C++ Standard Library whenever possible.
Prefer std::string over C-style strings.
Use std::vector, std::map, std::unordered_map, etc. for collections.
Use std::optional, std::variant, std::any for modern type safety.
Use std::filesystem for file operations.
Use std::chrono for time-related operations.
Concurrency
Use std::thread, std::mutex, std::lock_guard for thread safety.
Prefer task-based parallelism over thread-based parallelism.
Use std::atomic for atomic operations.
Avoid data races by proper synchronization.
Use thread-safe data structures when necessary.
Build a custom version for your project
Write your project details; to the end of Rule is added, and you get a file that is given to the AI without extra explanation. Everything is built in your browser and no data goes to the server.
Preview of the final file
Where do I put the file?
Claude Code.claude/skills/cpp-programming-guidelines/SKILL.md
Team projects: commit the file to the repository so every agent on the team follows the same rules. With Build a bundle Get several skills together, with the correct structure.
Frequently asked questions
Rule “cpp-programming-guidelines” — where do I put it?
for Claude Code in .claude/skills/cpp-programming-guidelines/SKILL.md, for Codex in .agents/skills, for Copilot in .github/skills and for Cursor in .cursor/skills or .cursor/rules. Use the “Build a bundle” button to get a zip with the correct structure.
Can I modify it for my own project?
Yes. The license CC0-1.0 allows modification and commercial use. With “Build a custom version” your project's details are added to the file, and you can edit the text afterward.
This Rule Which tools does it work with?
With Cursor. Standard Agent Skills are read by all modern coding agents.
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence…
Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems.…