Samples

Sample Apps

Complete, runnable sample applications demonstrating the a2a-swift SDK in real-world scenarios. Each sample is a standalone Swift package that you can build and run independently.

Looking for quick code snippets? Check out Examples/ for single-file reference implementations of agents and clients.

Apps

Sample	Description	Stack
A2AServer	Dockerized product catalog agent with Ollama LLM	Vapor · Docker · Ollama
A2AChatClient	macOS chat client with multi-agent support	SwiftUI · Foundation Models

Prerequisites

Before running the samples, make sure you have the following installed:

Tool	Required for	Install
Docker Desktop	A2AServer	brew install --cask docker
Xcode 26+	A2AChatClient	Mac App Store
Swift 6.0+	Both (local builds)	Included with Xcode

No Docker? You can also run the server locally with swift run — see A2AServer README for details.

Quick Start

Run both apps end-to-end in under 5 minutes:

1. Start the server

cd Samples/A2AServer
docker compose up --build

First run pulls the Ollama image and qwen3:0.6b model (~500MB). Wait until you see:

a2a-server-1  | Starting Product Catalog Agent (20 products) with Ollama at http://ollama:11434

2. Verify the server is running

# In a new terminal
curl http://localhost:8080/.well-known/agent-card.json | jq .name
# → "Tech Store Product Catalog"

3. Run the client

cd Samples/A2AChatClient
open Package.swift
# Xcode opens → Build & Run (⌘R)

4. Connect and chat

1. In the app sidebar, click "Connect an Agent"
2. Enter http://localhost:8080
3. Click Connect — you'll see the agent's name and skills
4. Try: "What laptops do you have?", then follow up with "Which one is cheaper?"

No Docker? Run the server locally

# Option A: With Ollama (natural language responses)
brew install ollama
ollama serve &
ollama pull qwen3:0.6b
cd Samples/A2AServer
OLLAMA_HOST=http://localhost:11434 swift run

# Option B: Without Ollama (returns raw search results, no setup needed)
cd Samples/A2AServer
swift run

Architecture

┌──────────────────────┐         A2A Protocol          ┌──────────────────────┐
│   A2AChatClient      │ ◄──── (JSON-RPC + SSE) ────► │   A2AServer          │
│                      │                               │                      │
│  SwiftUI + FM        │  1. Agent card discovery       │  Vapor + Ollama      │
│  orchestration       │  2. SendMessage (streaming)    │  product catalog     │
│                      │  3. Task continuations          │  conversation memory │
└──────────────────────┘                               └──────────────────────┘

Both apps use the a2a-swift SDK (A2A package) — the server uses AgentExecutor, DefaultRequestHandler, and A2ARouter; the client uses A2AClient for discovery and streaming communication.