convez.ai
MCP ServerIntegração & APIsIntermediário
Tom-Ryder

AgentRunKit

Tom-Ryder/AgentRunKit

SDK Swift para agents LLM type-safe com inferência local e em nuvem

Ver no GitHub
BomPluginScore · de 0 a 100

SDK Swift 6 type-safe para construir agents LLM com chamadas de ferramentas compiladas. Suporta inferência em nuvem (OpenAI) e local (MLX em Apple Silicon) com zero dependências no núcleo e full async/await.

Instalação

$npx convez-ai Tom-Ryder/AgentRunKit

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
38
Forks
8
Acompanhando
2
Saúde
57%
Último commit
há 1 mês
Versão
v6.1.0
Issues abertas
5
Contribuidores
3

Casos de uso

  • Construir agents autônomos em iOS/macOS com ferramentas type-safe
  • Integrar inferência LLM local em dispositivos Apple Silicon
  • Criar pipelines de coding agents com chamadas de ferramentas estruturadas
  • Implementar agents com streaming e controle de budget de tokens

Veredito

Ferramenta madura e bem-documentada para devs Swift que precisam de agents LLM com segurança de tipo. Oferece escolha real entre cloud e on-device, mas exige familiaridade com Swift 6 e async/await.

README

A tradução para o português deste README ainda está na fila.

AgentRunKit


Quick Start

import AgentRunKit

let client = OpenAIClient.openAI(apiKey: "sk-...", model: "gpt-5.4")

let weatherTool = try Tool<WeatherParams, String, EmptyContext>(
    name: "get_weather",
    description: "Get the current weather"
) { params, _ in
    "72°F and sunny in \(params.city)"
}

let agent = Agent(client: client, tools: [weatherTool])
let result = try await agent.run(userMessage: "What's the weather in SF?", context: EmptyContext())
if let content = result.content {
    print(content)
}

result.content is optional. Completed runs return finish-tool content, while structural terminal reasons such as max iterations or token budget exhaustion surface through result.finishReason with no final content.


Documentation

Full documentation including guides and API reference is available on Swift Package Index.


Runnable Example

Examples/AgentCode is an interactive terminal coding agent built with AgentRunKit. It demonstrates the full agent loop in a local workspace: streaming events, type-safe tools, approval-gated edits and command execution, bounded file access, transcript export, and deterministic offline mode.

cd Examples/AgentCode
swift run agent-code

By default it opens a bundled broken Swift package so you can ask it to fix failing tests. Set OPENAI_API_KEY for a live OpenAI-compatible provider, or run without a key to exercise the CLI with the offline test client.


Installation

Add to your Package.swift:

dependencies: [
    .package(url: "https://github.com/Tom-Ryder/AgentRunKit.git", from: "4.3.0")
]
.target(name: "YourApp", dependencies: ["AgentRunKit"])

For on-device inference, additional targets are available:

  • AgentRunKitMLX for MLX on Apple Silicon (links mlx-swift-lm)
  • AgentRunKitFoundationModels for Apple Foundation Models (iOS 26+ and macOS 26+, no external dependencies)

Features

  • Agent loop with configurable iteration limits and token budgets
  • Streaming with AsyncThrowingStream and @Observable SwiftUI wrapper
  • Type-safe tools with compile-time JSON schema validation
  • Sub-agent composition with depth control and streaming propagation
  • Context management: automatic compaction, pruning, token budgets
  • Checkpoint and resume: per-iteration snapshots, file or in-memory backends
  • Tool approval: human-in-the-loop policies with session allowlists
  • Structured output with JSON schema constraints
  • Multimodal input: images, audio, video, PDF
  • Text-to-speech with concurrent chunking and MP3 concatenation
  • MCP client: stdio transport, tool discovery, JSON-RPC
  • Extended thinking / reasoning model support

Providers

ProviderDescription
OpenAIClientOpenAI and compatible APIs (OpenRouter with reasoning replay, Groq, Together, Ollama)
AnthropicClientAnthropic Messages API with same-substrate continuity replay
GeminiClientGoogle Gemini API
VertexAnthropicClientAnthropic models on Google Vertex AI with same-substrate continuity replay
VertexGoogleClientGoogle models on Vertex AI
ResponsesAPIClientOpenAI Responses API with same-substrate continuity replay and OpenRouter factory
FoundationModelsClientApple on-device (macOS 26+ / iOS 26+)
MLXClientOn-device via MLX on Apple Silicon

Requirements

PlatformVersion
iOS18.0+
macOS15.0+
Swift6.0+
Xcode26+ for local development and CI

License

MIT License. See LICENSE for details.

Ver o original no GitHub

Bom para

Relacionados