rust-expert

name	rust-expert
description	Expert Rust idiomatique pour développement CLI/système. Ownership, error handling avec anyhow/thiserror, traits, async Tokio, testing. Utiliser pour coder, reviewer ou refactorer du Rust.
allowed-tools	Read, Grep, Glob, Bash
effort	medium
tags	["rust","expert","ownership","async","error-handling"]

Rust Expert Skill

Expert in idiomatic Rust 1.75+ development with focus on CLI/system tools, ownership patterns, error handling, traits, async programming, and testing.

Core Competencies

1. Ownership & Borrowing Mastery

• Prefer &str over String for function parameters
• Use slices (&[T]) over Vec<T> when ownership not needed
• Leverage Cow<str> for conditionally owned strings
• Apply Arc<T> and Arc<Mutex<T>> for thread-safe sharing
• Design lifetimes explicitly when necessary

2. Error Handling Excellence

• Default: anyhow::Result<T> for applications
• Libraries: thiserror for custom error types
• Context: Always use .context("operation description") with ?
• No unwrap: Use ? operator or expect() with justification
• Recovery: Provide actionable error messages

3. Trait-Driven Design

• Implement standard traits: Debug, Clone, Default, Display
• Use trait objects (dyn Trait) for runtime polymorphism
• Leverage trait bounds for generic constraints
• Organize impl blocks immediately after type definitions
• Apply async traits (native in Rust 1.75+)

4. Async Programming Patterns

• Runtime: Tokio for production async
• Patterns: async fn, tokio::spawn, tokio::select macro (with !)
• Streams: tokio_stream for async iteration
• Rate limiting: governor crate for API throttling
• Retries: Exponential backoff with tokio::time::sleep

5. CLI Development Best Practices

• Argument parsing: clap with derive macros
• Config: serde + toml/yaml for structured config
• Output: indicatif for progress, colored for styling
• Terminal: Handle SIGINT gracefully, cleanup on exit

6. Testing Strategy

• Embedded tests: #[cfg(test)] mod tests { use super::*; }
• Unit tests alongside code, integration tests in tests/
• Use #[should_panic] for expected panics
• Mock external dependencies with traits
• Property-based testing with proptest or quickcheck

7. Build Optimization

• Linker: Use mold or lld for faster linking
• Cache: sccache for incremental builds
• Profile: cargo build --release with LTO
• Dependencies: Minimize with feature flags
• Workspace: Organize large projects with workspace

8. Code Quality Standards

• Before commit: cargo fmt, cargo clippy, cargo test
• Clippy: Zero warnings policy
• Documentation: /// for public APIs, // with ! for modules
• Examples: Provide runnable examples in docs

9. Project Structure Patterns

• Flat module hierarchy when possible (rtk style)
• src/main.rs for binaries, src/lib.rs for libraries
• Group related functionality in modules
• Keep functions focused and small
• Avoid deep nesting (max 3-4 levels)

Reference Files

This skill includes detailed patterns and checklists:

• Patterns:

  • patterns/error-handling.md - anyhow, thiserror, Result patterns
  • patterns/ownership.md - &str vs String, Cow, Arc, lifetimes
  • patterns/async-patterns.md - Tokio, futures, async traits
  • patterns/traits-impl.md - Trait implementation organization
  • patterns/testing.md - Test organization and assertions
  • patterns/cli-patterns.md - clap derive, subcommands

• Checklists:

  • checklists/code-review.md - Pre-commit review checklist
  • checklists/performance.md - Build optimization checklist

• Anti-Patterns:

  • anti-patterns/common-mistakes.md - Frequent Rust mistakes
  • anti-patterns/memory-pitfalls.md - Ownership pitfalls

• Examples:

  • examples/rtk-patterns.md - Real-world patterns from rtk project

When to Use This Skill

✅ Use for:

• Writing new Rust code (CLI tools, libraries, async services)
• Code review of Rust PRs
• Refactoring to idiomatic patterns
• Performance optimization
• Error handling improvements
• Test organization

❌ Don't use for:

• Simple syntax questions (use native knowledge)
• Non-Rust languages
• Project setup without coding

Integration with SuperClaude Framework

• Works with: backend-architect (system design), TDD (test-first), code-reviewer
• Flags: Responds to --think for architectural analysis
• Mode: Compatible with --uc token efficiency mode
• Language: English for shareability, French instructions supported

Quick Reference

Error Handling

use anyhow::{Context, Result};

fn process_file(path: &str) -> Result<()> {
    let content = std::fs::read_to_string(path)
        .context("Failed to read file")?;
    Ok(())
}

Ownership Patterns

// Prefer borrowing
fn print_name(name: &str) { println!("{}", name); }

// Conditionally owned
use std::borrow::Cow;
fn maybe_modify(input: &str, modify: bool) -> Cow<str> {
    if modify {
        Cow::Owned(input.to_uppercase())
    } else {
        Cow::Borrowed(input)
    }
}

Async Patterns

#[tokio::main]
async fn main() -> Result<()> {
    let result = tokio::time::timeout(
        Duration::from_secs(5),
        fetch_data()
    ).await??;
    Ok(())
}

Testing

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_calculation() {
        assert_eq!(calculate(2, 3), 5);
    }
}