convez.ai
MCP ServerDevOps & InfraestruturaAvançado
ttoss

SOAT — Infraestrutura para Agentes IA em Produção

ttoss/soat

Infraestrutura self-hostable completa para agentes IA com memória, armazenamento e orquestração

Ver no GitHub
InicialPluginScore · de 0 a 100

Framework open-source que fornece uma camada de infraestrutura completa para aplicações de IA, incluindo gerenciamento de identidade, armazenamento de documentos com busca semântica, memória persistente para agentes, orquestração multi-agente e um servidor MCP nativo. Expõe todas as operações via REST, MCP, CLI e SDK TypeScript, eliminando a necessidade de reconstruir infraestrutura em cada projeto.

Instalação

$npx convez-ai ttoss/soat

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
10
Forks
2
Acompanhando
0
Saúde
37%
Último commit
há 12 horas
Versão
v0.6.0
Issues abertas
14
Contribuidores
3

Casos de uso

  • Construir aplicações de IA com persistência de memória para agentes autônomos
  • Implementar workflows multi-agente com rastreamento e observabilidade
  • Integrar agentes IA no Claude Desktop ou ambientes compatíveis com MCP
  • Gerenciar identidade, chaves API e segredos criptografados em projetos de IA
  • Recuperação aumentada por geração (RAG) com busca semântica em documentos

Veredito

SOAT é uma solução robusta e completa para infraestrutura de agentes IA em produção, com suporte nativo a MCP e múltiplas superfícies de acesso. Ideal para equipes que precisam de um backend escalável e self-hostable, embora exija conhecimento técnico avançado em DevOps e TypeScript para implementação.

README

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

SOAT — Infrastructure for production-ready AI agents

License: MIT SafeSkill 92/100

SOAT is open-source infrastructure for building AI applications. One self-hostable Node.js server gives you IAM, file and document storage with vector search, conversational memory, agent orchestration, multi-agent workflows, retrieval-augmented generation, declarative stack deployment, and a full Model Context Protocol server — backed by PostgreSQL.

You bring the product. SOAT handles the infrastructure layer.

Why SOAT?

Shipping AI applications means rebuilding the same infrastructure on every project: users, API keys, encrypted secrets, file storage, embeddings, conversation history, agent tool calling, traces, observability. SOAT solves all of it once and exposes it through four equivalent surfaces — REST, MCP, CLI, and TypeScript SDK — so the same operation runs the same way whether you call it from a backend, Claude Desktop, a CI script, or your own UI.

Highlights

  • Identity & access — users, projects, JWTs, project keys, personal API keys, and reusable IAM policy documents enforced consistently on every surface.
  • Files, documents, memories & knowledge — pgvector-backed semantic search across documents and memory entries, with memory containers that keep durable context for agents. See Files, Documents, Memories, and Knowledge.
  • Agents that actually do things — tool-calling reasoning loops with HTTP, MCP, client-side, and SOAT-platform tools, including multi-agent workflows where agents invoke other agents. See Agents.
  • Declarative deployments — Agent Formations let you define providers, memories, tools, and agents in a single template and deploy in dependency order.
  • Sessions — a 1-user ↔ 1-agent interface that hides actors and conversations. Two API calls take a user from message to answer. See Sessions.
  • Async generations — kick off long-running agent runs, poll for status, or fire a webhook on completion.
  • Operations — encrypted secrets, HMAC-signed webhooks with event-pattern subscriptions, and trace records for every generation. See Webhooks and Traces.
  • MCP native — every operation is automatically available as an MCP tool. Plug SOAT into Claude Desktop, Cursor, or any MCP-compatible runtime. See MCP docs.

Documentation

Read the full documentation — quick start, key concepts, module reference, API, MCP, CLI, and SDK guides.

Getting Started

The fastest path is Docker Compose:

git clone https://github.com/ttoss/soat.git
cd soat

Follow the Getting Started Guide to bring up the server and database in under five minutes.

Client surfaces

SurfaceBest forDocs
REST APIBackend services and custom integrationshttps://soat.ttoss.dev/docs/api
MCP serverClaude Desktop, Cursor, and other MCP-aware AI runtimeshttps://soat.ttoss.dev/docs/mcp
CLI (soat)Scripts, CI pipelines, and local explorationhttps://soat.ttoss.dev/docs/cli
SDK (@soat/sdk)TypeScript and JavaScript applicationshttps://soat.ttoss.dev/docs/sdk

All four hit the same business logic and the same permission engine.

Repository layout

This is a monorepo managed with pnpm and Turbo:

  • packages/server — the SOAT server (REST + MCP, business logic, permissions)
  • packages/postgresdb — Sequelize models and database utilities
  • packages/sdk — TypeScript SDK generated from the OpenAPI specs
  • packages/cli — the soat command-line client
  • packages/website — the documentation site (Docusaurus)

Contributing

Issues and pull requests are welcome. Run the tests with pnpm --filter @soat/server test and the smoke suite with pnpm run -w smoke-tests.

License

MIT — see the LICENSE file for details.

Acknowledgments

  • ttoss — for @ttoss/http-server and @ttoss/http-server-mcp
  • pgvector — PostgreSQL vector similarity search
  • Ollama — local LLM and embedding models
Ver o original no GitHub

Bom para

Relacionados