MCP e chamadas de ferramentas

OpenConnector

oomol-lab/open-connector

An open-source authentication gateway that connects over 1,000 SaaS providers to AI agents through SDK, CLI, MCP, HTTP, and OpenAPI interfaces.

Abrir repositório oficial ↗ Página inicial do projeto ↗
★ 4,1KEstrelas
⑂ 311Forks
4Problemas em aberto
TypeScriptIdioma
Apache-2.0Licença
Q@project.QualityScorePontuação editorial

Capturas de tela do projeto

Captura de tela de OpenConnector Captura de tela de OpenConnector Captura de tela de OpenConnector Captura de tela de OpenConnector Captura de tela de OpenConnector Captura de tela de OpenConnector

Visão geral

OpenConnector is an open-source connector gateway designed to link AI agents with external SaaS applications. Positioned as an alternative to Pipedream and Composio, it allows developers to connect user app accounts once and subsequently expose a shared catalog of over 1,000 providers and 10,000 prebuilt actions to agents and applications. The platform maintains credentials, scopes, schemas, policies, and run logs inside an inspectable runtime, ensuring that provider secrets remain behind a secure boundary and are never directly exposed to the agent process. It supports multiple deployment options, including local Docker or Node.js environments, Cloudflare Workers, Fly.io, and a managed OOMOL hosted runtime.

Principais recursos

  • Supports SDK, CLI, MCP, HTTP, and OpenAPI interfaces for agent connectivity
  • Catalog of 1,000+ providers and 10,000+ prebuilt actions
  • Inspectable action contracts including request/response schemas and required scopes
  • Credential handling for API keys, OAuth2, custom credentials, and no-auth providers
  • Runtime controls for connection identity, scopes, tokens, and redacted run logs
  • Multiple deployment options: local Docker/Node.js, Cloudflare Workers, Fly.io, and OOMOL hosted
  • Local dashboard for browsing connectors, configuring credentials, and monitoring usage

Requisitos, instalação e início rápido

For a self-hosted runtime using Docker Compose, run: `docker compose up`. This pulls the `ghcr.io/oomol-lab/open-connector:latest` image. To build from source instead, run: `docker compose -f docker-compose.yml -f docker-compose.build.yml up --build`. For npm-based local development, use Node.js 22+ and run `npm install` followed by `npm run dev`.

Uso

After starting the runtime, access the local console at `http://localhost:3000` and the API reference at `http://localhost:3000/docs`. To verify the runtime, run a no-auth action via curl: `curl -s -X POST http://localhost:3000/v1/actions/hackernews.get_top_stories -H 'content-type: application/' -d '{"input":{}}'`. To connect a provider like GitHub using a personal access token: `curl -s -X PUT http://localhost:3000/api/connections/github -H 'content-type: application/' -d '{"authType":"api_key","values":{"apiKey":"github_pat_..."}}'`.

Compatibilidade do modelo e casos de uso

Not stated in the repository metadata

Observações sobre licença e riscos

The source code, scripts, generated project scaffolding, tests, and documentation are licensed under the Apache License, Version 2.0. The license does not grant rights to third-party products, providers, apps, APIs, trademarks, or brand assets. Provider and app names are included solely for identification and interoperability.

Editorial verification 2026-08-09: repository URL, owner, description, license and repository statistics were reviewed. License metadata: Apache-2.0. README was fetched for the channel draft; re-check repository dependencies, releases and model terms before production use.

Lançamento e manutenção

Not stated in the repository metadata

MemPalace

mempalace/mempalace

★ 58KPython

QwenPaw

agentscope-ai/qwenpaw

★ 33,9KPython

Composio

composiohq/composio

★ 29,5KTypeScript