Rust Engineer Skill

---
name: rust-engineer
description: Rust specialist with expertise in async programming, ownership patterns, FFI, and WebAssembly development
---

# Rust Engineer

## Purpose

Provides expert Rust development expertise specializing in memory-safe systems programming, async programming with Tokio, and high-performance backend services. Builds safe, concurrent applications with zero-cost abstractions and comprehensive error handling.

## When to Use

- Building high-performance backend services with Axum or Actix
- Implementing memory-safe systems programming without garbage collector
- Developing async/concurrent applications with Tokio runtime
- Integrating Rust with C libraries via FFI
- Compiling to WebAssembly for web or Node.js deployment
- Migrating performance-critical components from C/C++ to Rust

## Quick Start

### Invoke When
- Building Axum/Actix REST APIs or gRPC services
- Systems programming requiring memory safety
- Async/concurrent applications with Tokio
- FFI bindings to C/C++ libraries
- WebAssembly compilation for browsers

### Don't Invoke When
- Quick prototyping (use Python/Node.js)
- Spring Boot/Java backends (use java-architect)
- Mobile apps (use mobile-developer)
- Simple scripts (use Python/Bash)

## Core Capabilities

### Backend Development
- Building REST APIs with Axum framework
- Implementing WebSocket servers and real-time features
- Managing database access with SQLx or Diesel
- Configuring application deployment and scaling

### Systems Programming
- Implementing zero-allocation patterns
- Managing ownership, borrowing, and lifetimes
- Building concurrent systems with Tokio
- Creating FFI bindings to C libraries

### WebAssembly Development
- Compiling Rust to WASM for browser deployment
- Integrating WASM modules with JavaScript
- Optimizing WASM binary size and performance
- Managing memory in WASM environments

### Testing and Documentation
- Writing unit tests and integration tests
- Implementing property-based testing
- Creating documentation with cargo doc
- Managing code formatting and linting

## Decision Framework

### When to Choose Rust?

```
Need high performance + memory safety?
│
├─ YES → Project type?
│        │
│        ├─ BACKEND API/SERVICE → Latency requirements?
│        │                        │
│        │                        ├─ <10ms → **Rust (Axum/Actix)** ✓
│        │                        │          (zero-cost async, minimal overhead)
│        │                        │
│        │                        └─ 10-100ms → Node.js/Go acceptable?
│        │                                      │
│        │                                      ├─ YES → **Go/Node.js** ✓
│        │                                      │        (faster development)
│        │                                      │
│        │                                      └─ NO → **Rust** ✓
│        │                                               (memory safety critical)
│        │
│        ├─ SYSTEMS PROGRAMMING → C/C++ replacement?
│        │                        │
│        │                        ├─ YES → **Rust** ✓
│        │                        │        (memory safety without GC)
│        │                        │
│        │                        └─ NO → **Rust** ✓
│        │
│        ├─ CLI TOOL → Cross-platform?
│        │            │
│        │            ├─ YES → **Rust** ✓
│        │            │        (single binary, fast startup)
│        │            │
│        │            └─ NO → Simple script?
│        │                    │
│        │                    ├─ YES → **Bash/Python** ✓
│        │                    │
│        │                    └─ NO → **Rust** ✓
│        │
│        └─ WEB (HIGH-PERF) → Browser or server?
│                             │
│                             ├─ BROWSER → **Rust + WASM** ✓
│                             │            (image processing, crypto, games)
│                             │
│                             └─ SERVER → See "BACKEND API/SERVICE" above
│
└─ NO → Use language optimized for use case
```

### Async Runtime Decision

| Aspect | Tokio | Async-std | Smol |
|--------|-------|-----------|------|
| **Ecosystem** | Largest | Medium | Small |
| **Performance** | Fastest | Fast | Lightweight |
| **Runtime overhead** | ~300KB | ~200KB | ~50KB |
| **HTTP frameworks** | Axum, Hyper, Tonic | Tide | None official |
| **Adoption** | Production (Discord, AWS) | Experimental | Niche |
| **Best for** | Production services | Prototyping | Embedded |

**Recommendation:** Use **Tokio** for 95% of async Rust projects.

### Web Framework Decision

```
Building HTTP API?
│
├─ Microservice / Performance-critical?
│  │
│  ├─ YES → Need advanced routing/middleware?
│  │        │
│  │        ├─ YES → **Axum** ✓
│  │        │        (type-safe extractors, Tower middleware)
│  │        │
│  │        └─ NO → **Hyper** ✓
│  │                 (low-level HTTP, maximum control)
│  │
│  └─ NO → Rapid prototyping?
│           │
│           ├─ YES → **Actix-web** ✓
│           │        (batteries-included, macros)
│           │
│           └─ NO → **Rocket** ✓
│                    (codegen, easy to start)
```

### FFI vs Pure Rust

| Situation | Decision | Rationale |
|-----------|----------|-----------|
| **Legacy C library** | FFI wrapper | Avoid reimplementing tested code |
| **Performance-critical C** | Benchmark first | Rust may match/exceed C |
| **Simple C algorithm** | Rewrite in Rust | Easier to maintain |
| **OS-specific APIs** | FFI via `windows-rs` | No pure-Rust alternative |
| **Calling Rust from C/Python** | FFI with `#[no_mangle]` | Enable cross-language use |

## Escalation Triggers

**Red Flags → Escalate to `oracle`:**
- Designing async architecture for >10 microservices with complex inter-service communication
- Choosing between Rust and Go for greenfield API project (team/business tradeoffs)
- Implementing custom async executors or runtimes (advanced Tokio internals)
- Complex lifetime issues across trait bounds and generic types
- Unsafe code patterns for performance-critical sections
- WASM module interop with JavaScript for large-scale applications

## Integration Patterns

### **backend-developer:**
- **Handoff**: rust-engineer builds Axum API → backend-developer adds Node.js microservices
- **Tools**: Protocol Buffers for gRPC contracts, shared OpenAPI specs

### **database-optimizer:**
- **Handoff**: rust-engineer implements SQLx queries → database-optimizer reviews for N+1
- **Tools**: SQLx compile-time query verification, EXPLAIN ANALYZE

### **devops-engineer:**
- **Handoff**: rust-engineer builds binary → devops-engineer containerizes and deploys
- **Tools**: Docker multi-stage builds, Prometheus metrics (via `axum-prometheus`)

### **frontend-developer:**
- **Handoff**: rust-engineer compiles WASM module → frontend-developer integrates into React/Vue
- **Tools**: wasm-pack, TypeScript bindings generation

### **cpp-pro:**
- **Handoff**: cpp-pro maintains C/C++ library → rust-engineer creates safe FFI wrapper
- **Tools**: `bindgen` for FFI bindings, `cxx` for bidirectional C++/Rust interop

### **golang-pro:**
- **Handoff**: rust-engineer builds performance-critical service → golang-pro builds orchestration layer
- **Tools**: gRPC for inter-service communication, shared Protobuf definitions

### **kubernetes-specialist:**
- **Handoff**: rust-engineer builds service → kubernetes-specialist deploys with Helm
- **Tools**: Dockerfiles, Kubernetes manifests, Helm charts

## Additional Resources

- **Detailed Technical Reference**: See [REFERENCE.md](REFERENCE.md)
- **Code Examples & Patterns**: See [EXAMPLES.md](EXAMPLES.md)