julia — скилл для Claude Code: правила разработки на Julia

---
name: julia
description: Julia development guidelines covering multiple dispatch, type system, performance optimization, and scientific computing best practices.
metadata:
  maintainer: Mindrally
  source: https://github.com/Mindrally/skills
---

# Julia Development

You are an expert in Julia programming with deep knowledge of multiple dispatch, the type system, and high-performance computing.

## Core Principles

- Write concise, technical responses with accurate Julia examples
- Leverage multiple dispatch and the type system for performant code
- Prefer immutable structs and functions over mutable state
- Use Julia's built-in features for parallelism and performance

## Naming Conventions

- **Functions/variables**: snake_case (e.g., `process_data`, `is_active`)
- **Types**: PascalCase for structs and abstract types
- **Files/directories**: lowercase with underscores (e.g., `src/data_processing.jl`)

## Function Guidelines

All functions require docstrings with signatures and return value descriptions:

```julia
"""
    process_data(data::Vector{Float64}, threshold::Float64) -> Vector{Float64}

Process input data by applying a threshold filter.
"""
function process_data(data::Vector{Float64}, threshold::Float64)
    # implementation
end
```

## Struct Definitions

- Use `@kwdef` macro for keyword constructors
- Include comprehensive docstrings for each field
- Implement custom `show` methods using `dump`
- Prefer immutable structs unless mutation is required

## Error Handling

- Create custom exception types for domain-specific errors
- Use guard clauses for preconditions
- Example: `x <= 0 && throw(InvalidInputError("Input must be positive"))`
- Provide informative error messages

## Performance Optimization

- Use type annotations to prevent type instability
- Prefer statically sized arrays (SArray) for fixed collections
- Use `@views` macro to avoid unnecessary copying
- Leverage built-in parallelism with `@threads` and `@distributed`
- Profile with BenchmarkTools.jl before optimizing
- Avoid global variables in performance-critical code

## Testing Structure

- Use the `Test` module with one top-level `@testset` per file
- Individual `@test` calls assess basic functionality
- Test edge cases and type stability separately
- Use `@test_throws` for expected errors

## Code Organization

- Organize functionality through modules
- Use abstract types with multiple dispatch for separation
- Maintain consistent project structure (src/, test/, docs/)
- Export only public API functions
- Use `include` for organizing large modules