Agent-Skills.md

Agent Skills: SOLID Rust - Modular Architecture

SOLID principles for Rust 2024+. Files < 100 lines, traits separated, modular architecture. Modules MANDATORY.

UncategorizedID: fusengine/agents/solid-rust

Repository

fusengine/agents
fusengineLicense: MIT
7

Install this agent skill to your local

pnpm dlx add-skill https://github.com/fusengine/agents/tree/HEAD/plugins/solid/skills/solid-rust

Skill Files

Browse the full folder contents for solid-rust.

Download Skill

Loading file tree…

plugins/solid/skills/solid-rust/SKILL.md

Skill Metadata

Name
solid-rust
Description
"Use when writing or refactoring Rust code, structuring src/modules/[feature]/ layers, or defining traits.rs interfaces (SOLID, files within FUSE_SOLID_MAX_LINES)."
<objective> SOLID Rust enforces a modular architecture for Rust 2024 edition: every feature lives under `src/modules/[feature]/` (handlers.rs, services.rs, repository.rs, traits.rs, models.rs) with shared code in `src/core/`, traits live only in `traits.rs` or `src/core/traits/`, generics use trait bounds instead of concrete types, `thiserror` handles custom errors, and every public item carries a rustdoc `///` comment.

Before writing any new code it requires a DRY check against src/core/services and src/core/traits. See solid-principles.md for the overview, the per-principle references for SRP/OCP/LSP/ISP/DIP detail, and the templates for module/service/trait/handler/error/test scaffolding. </objective>

SOLID Rust - Modular Architecture

Agent Workflow (MANDATORY)

Before ANY implementation, spawn 3 agents in parallel, one Agent call each with a name:

  1. fuse-ai-pilot:explore-codebase - Analyze existing architecture
  2. fuse-ai-pilot:research-expert - Verify Rust docs via Context7
  3. fuse-ai-pilot:sniper - Post-implementation validation

DRY - Reuse Before Creating (MANDATORY)

Before writing ANY new code:

  1. Grep the codebase for similar traits, services, or logic
  2. Check shared locations: src/core/services/, src/core/traits/
  3. If similar code exists -> extend/reuse instead of duplicate
  4. If code will be used by 2+ features -> create it in src/core/

Architecture (Modules MANDATORY)

| Layer | Location | |-------|----------| | Handlers | src/modules/[feature]/handlers.rs | | Services | src/modules/[feature]/services.rs | | Repositories | src/modules/[feature]/repository.rs | | Traits | src/modules/[feature]/traits.rs | | Models | src/modules/[feature]/models.rs | | Shared | src/core/{services,traits,models}/ |

NEVER use flat src/ structure - always src/modules/[feature]/

Critical Rules (MANDATORY)

| Rule | Value | |------|-------| | File limit | FUSE_SOLID_MAX_LINES | | Handlers | Thin, delegate to services | | Traits | traits.rs or src/core/traits/ ONLY | | Rustdoc | /// on every public item | | Error handling | Use thiserror for custom errors | | Generics | Use trait bounds, not concrete types |

Reference Guide

Concepts

| Topic | Reference | When to consult | |-------|-----------|-----------------| | SOLID Overview | solid-principles.md | Quick reference | | SRP | single-responsibility.md | Fat structs | | OCP | open-closed.md | Adding impls | | LSP | liskov-substitution.md | Trait contracts | | ISP | interface-segregation.md | Fat traits | | DIP | dependency-inversion.md | Generics/DI | | Architecture | architecture-patterns.md | Modular crate |

Templates

| Template | When to use | |----------|-------------| | module.md | Feature module structure | | service.md | Business logic service | | trait-def.md | Trait definition | | handler.md | HTTP handler (Axum) | | error.md | Custom errors (thiserror) | | test.md | Unit + integration tests |

Forbidden

| Anti-Pattern | Fix | |--------------|-----| | Files over FUSE_SOLID_MAX_LINES | Split by responsibility | | Traits in impl files | Move to traits.rs | | Box<dyn Any> | Use proper trait bounds | | Flat src/ structure | Use src/modules/[feature]/ | | Unwrap in library code | Use Result<T, E> |