agentsclimarketplace

Server side swift

Skill Insolvent-pyrene754/claude-code-server-side-swift-skills/skills/server-side-swift

Server-side Swift development with Vapor and Hummingbird. Covers project setup, routing, Fluent ORM, authentication, deployment, and shared client-server code. Use when building backends or APIs in Swift.From its SKILL.md

Install
npx -y skills add Insolvent-pyrene754/claude-code-server-side-swift-skills --skill server-side-swift

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

9.3 KB, ~2.2k tokens by cl100k_base, as published. Nobody here has run it

Server-Side Swift Generator

Generates and guides server-side Swift projects using Vapor or Hummingbird, including API design, database integration, authentication, deployment, and shared code with iOS clients.

When This Skill Activates

Use this skill when the user:

  • Asks to "create a backend" or "build an API" in Swift
  • Mentions "Vapor", "Hummingbird", or "server-side Swift"
  • Wants JWT, OAuth, or session authentication on the server
  • Needs to share Codable models between an iOS app and a Swift server
  • Asks about sending push notifications from a server (APNSwift)
  • Wants to deploy a Swift server (Docker, Fly.io, Railway)
  • Mentions Fluent ORM, database migrations, or PostgreSQL with Swift
  • Asks about WebSockets or real-time features on the server
  • Wants to replace CloudKit with a custom backend

Pre-Generation Checks

1. Project Context Detection

  • Check if a Package.swift already exists
  • Determine if this is a new project or adding to an existing one
  • Check Swift version (async/await requires Swift 5.5+, Hummingbird 2 requires Swift 6+)
  • Search for existing server-side code
Glob: **/Package.swift, **/configure.swift, **/routes.swift, **/entrypoint.swift
Grep: "import Vapor" or "import Hummingbird" or "import Fluent"

2. Conflict Detection

Search for existing server implementations:

Glob: **/Controllers/*.swift, **/Migrations/*.swift, **/Models/*.swift
Grep: "RouteCollection" or "AsyncMigration" or "func routes"

If found, ask user:

  • Extend existing server with new endpoints?
  • Replace existing implementation?

3. iOS Client Detection

Check if an iOS client exists in the workspace:

Glob: **/*.xcodeproj, **/*.xcworkspace
Grep: "import SwiftUI" or "import UIKit"

If found, recommend SharedModels package pattern.

Architecture Decision Gate

Before asking about framework choice, validate that server-side Swift is the right technology for this project. Reference architecture-decisions.md for the full decision framework.

Quick Validation (ask yourself, not the user)

  1. Does the user have a Swift team or iOS app? → If no, surface alternatives (Go, Node.js, Python)
  2. Does client-server code sharing provide value? → Strongest signal for Swift
  3. Does the project need integrations Swift lacks (Kafka, Elasticsearch, email)? → Surface ecosystem gaps honestly

If any red flags, present trade-offs from architecture-decisions.md before proceeding. Example:

"Before we start — server-side Swift works best when you have an iOS app that can share models with the backend. Since your project is [web-only / cross-platform / ML-heavy], you might want to consider [Node.js / Go / Python] instead. Want me to explain the trade-offs, or proceed with Swift?"

If "Help me decide" on Vapor vs Hummingbird

Don't default arbitrarily. Ask about context:

  • Team size and experience — New to server Swift? → Vapor (more docs, bigger community)
  • Project type — Lambda/microservice? → Hummingbird (smaller binary, first-class Lambda)
  • Compile time sensitivity — Tight CI/CD? → Hummingbird (lighter dependency graph)
  • Feature needs — Need ORM + auth + WebSocket out of the box? → Vapor

Reference the weighted decision matrix in architecture-decisions.md for detailed comparison.

Configuration Questions

Ask user via AskUserQuestion:

  1. Which framework do you want to use?

    • Vapor (batteries-included, Fluent ORM, larger community — 78% market share)
    • Hummingbird (lightweight, pure structured concurrency, modular — ideal for microservices/Lambda)
    • Help me decide (I'll ask about your project context)
  2. What database do you need?

    • PostgreSQL (recommended for production)
    • SQLite (development/lightweight)
    • None (stateless API)
    • MongoDB
  3. What authentication method?

    • JWT (stateless, recommended for mobile APIs)
    • Session-based (web apps)
    • Sign in with Apple (server-side validation)
    • Bearer token (simple)
    • None
  4. Do you need shared models with an iOS client?

    • Yes, create a SharedModels package
    • No, server-only project

Generation Process

Step 1: Project Scaffolding

Generate the project structure based on framework choice:

Vapor:

Sources/App/
├── Controllers/
├── Migrations/
├── Models/
├── DTOs/
├── configure.swift
├── entrypoint.swift
└── routes.swift
Tests/AppTests/
Package.swift
Dockerfile
docker-compose.yml
.env
.gitignore

Hummingbird:

Sources/App/
├── Controllers/
├── Models/
├── Application+build.swift
└── App.swift
Tests/AppTests/
Package.swift
Dockerfile

Step 2: Core Files

Generate based on configuration:

  1. Package.swift - Dependencies for chosen framework + database + auth
  2. configure.swift / Application+build.swift - App configuration
  3. Entry point with environment detection
  4. Basic health check endpoint

Step 3: Database Layer (if selected)

  1. Model files with Fluent property wrappers
  2. Migration files with AsyncMigration
  3. Database configuration in configure.swift

Step 4: Authentication (if selected)

  1. Auth middleware and authenticators
  2. Token/session model and migration
  3. Protected route groups
  4. Login/register endpoints

Step 5: Shared Models (if selected)

  1. SharedModels/ package with DTOs
  2. Update both Package.swift files to depend on SharedModels
  3. API endpoint protocol definitions

Step 6: Deployment Files

  1. Multi-stage Dockerfile
  2. docker-compose.yml with database service
  3. .env template with required variables

Swift 6 Concurrency Notes

Vapor 4 Async/Await

All new code should use async/await exclusively:

app.get("users") { req async throws -> [User] in
    try await User.query(on: req.db).all()
}

Hummingbird 2 Pure Concurrency

Hummingbird 2 is built entirely on structured concurrency with no EventLoopFuture APIs:

router.get("users") { request, context async throws -> [UserDTO] in
    try await userRepository.findAll()
}

Sendable Compliance

All DTOs shared between client and server must be Sendable:

public struct UserDTO: Codable, Sendable {
    public let id: UUID?
    public let name: String
    public let email: String
}

Output Format

Files Created

YourProject/
├── Sources/App/
│   ├── Controllers/
│   │   └── TodoController.swift    # Route collection
│   ├── Migrations/
│   │   └── CreateTodo.swift        # Database migration
│   ├── Models/
│   │   └── Todo.swift              # Fluent model
│   ├── DTOs/
│   │   └── TodoDTO.swift           # Request/response DTO
│   ├── configure.swift             # App configuration
│   ├── entrypoint.swift            # Entry point
│   └── routes.swift                # Route registration
├── Tests/AppTests/
│   └── TodoTests.swift             # Integration tests
├── Package.swift                   # Dependencies
├── Dockerfile                      # Multi-stage build
├── docker-compose.yml              # App + database
└── .env                            # Environment variables

Integration Steps

Run locally:

# Start database
docker-compose up db -d

# Run migrations
swift run App migrate --yes

# Start server
swift run App serve --hostname 0.0.0.0 --port 8080

Connect from iOS client:

let client = URLSessionAPIClient(
    configuration: .init(baseURL: URL(string: "http://localhost:8080/api/v1")!)
)
let todos = try await client.request(ListTodosEndpoint())

Testing

@Suite("API Tests", .serialized)
struct TodoTests {
    @Test("Create todo returns 200")
    func createTodo() async throws {
        try await withApp(configure: configure) { app in
            let dto = TodoDTO(title: "Test")
            try await app.testing().test(.POST, "api/v1/todos", beforeRequest: { req in
                try req.content.encode(dto)
            }, afterResponse: { res async throws in
                #expect(res.status == .ok)
                let created = try res.content.decode(TodoDTO.self)
                #expect(created.title == "Test")
            })
        }
    }
}

References

  • architecture-decisions.md - When to use server-side Swift, Vapor vs Hummingbird decision matrix, ecosystem gaps, pitfalls
  • vapor-patterns.md - Routing, controllers, middleware, configuration
  • hummingbird-patterns.md - Hummingbird 2 patterns and Vapor vs Hummingbird decision
  • database-patterns.md - Fluent ORM, migrations, queries, relationships
  • auth-patterns.md - JWT, sessions, Bearer, Sign in with Apple server-side
  • deployment-patterns.md - Docker, Fly.io, Railway, production setup
  • shared-code-patterns.md - SharedModels, DTOs, APNSwift
  • testing-patterns.md - VaporTesting, Swift Testing, integration tests
  • Vapor Documentation
  • Hummingbird Documentation
  • Swift on Server

What ships with it: 9 files

564.6 KB alongside SKILL.md

Gives 0 of the 12 instructions most data backend skills give in ~2.2k tokens

Counted across 229 of the 229 authors here whose files we hold, read 2026-08-07

  • Separate business logic into service layersin 22 of 229, across 15 files
  • Retry failures with exponential backoffin 21 of 229, across 14 files
  • Select only needed database columnsin 20 of 229, across 13 files
  • Abstract data access into repository classesin 19 of 229, across 12 files
  • Use centralized error handlersin 17 of 229, across 10 files
  • Use AsNoTracking for read-only queriesin 16 of 229, across 4 files
  • Use async/await for all I/O operationsin 16 of 229, across 5 files
  • Implement structured loggingin 15 of 229, across 4 files
  • Use dependency injection for all servicesin 14 of 229, across 2 files
  • Use resource-based URLs for REST APIsin 13 of 229, across 7 files
  • Invalidate cache after data changesin 13 of 229, across 9 files
  • Use a dependency injection containerin 12 of 229, across 4 files

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 326,871. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.