convez.ai
MCP ServerProdutividadeIntermediário
Zayrick

UniLLMs

Zayrick/UniLLMs

Cliente iOS nativo para gerenciar múltiplos LLMs e MCP servers

Ver no GitHub
InicialPluginScore · de 0 a 100

Cliente nativo iOS para chat com LLMs que gerencia múltiplos provedores (OpenAI, Anthropic, Gemini, etc.), conversas, prompts, memórias, ferramentas e servidores MCP em um único app. Armazena tudo localmente com foco em privacidade.

Instalação

$npx convez-ai Zayrick/UniLLMs

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
0
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 1 mês
Versão
—
Issues abertas
0
Contribuidores
1

Casos de uso

  • Usar múltiplos provedores de IA sem alternar apps
  • Gerenciar conversas com contexto persistente e memórias de longo prazo
  • Conectar ferramentas externas via MCP servers diretamente no cliente
  • Manter privacidade com armazenamento local de APIs e configurações

Veredito

Ferramenta especializada e bem arquitetada para quem trabalha com múltiplos LLMs em iOS e precisa de controle total via MCP servers. Licença AGPLv3 e privacidade por design são destaques, mas requer conhecimento técnico para compilar e integrar.

README

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

UniLLMs

UniLLMs is a native iOS LLM chat client for configuring multiple model providers and managing conversations, system prompts, memories, tools, MCP servers, and attachments in one app.

Features

  • Configure OpenRouter, OpenAI, Anthropic, Gemini, Pollinations, and OpenAI-compatible providers.
  • Manage models, system prompts, long-term memories, and tool settings.
  • Display chat responses as plain text with attachment previews.
  • Configure MCP servers to connect external tools to the chat runtime.
  • Access permission status, project information, contact details, source repository, and privacy policy from Settings.

Privacy

UniLLMs does not operate an analytics service and does not collect personal information, usage telemetry, advertising identifiers, crash reports, or diagnostic logs from your device.

Conversations, provider configurations, API keys, model metadata, memories, system prompts, tool settings, MCP server settings, and app preferences are stored locally. When you send a message, the app sends the conversation content and context required to complete the request to the provider you configured. If you enable MCP servers, custom API bases, or third-party tools, related requests are also sent to the endpoints you configured. Review those services' terms, logging behavior, retention policies, and privacy policies before using them.

Development

Open the Xcode project:

open UniLLMs.xcodeproj

List available schemes:

xcodebuild -list -project UniLLMs.xcodeproj

Compile the iOS Simulator app target without launching the app:

xcodebuild -project UniLLMs.xcodeproj \
  -scheme UniLLMs \
  -configuration Debug \
  -sdk iphonesimulator \
  -destination 'generic/platform=iOS Simulator' \
  -derivedDataPath .DerivedData \
  build

Compile the test bundle without running tests:

xcodebuild build-for-testing \
  -project UniLLMs.xcodeproj \
  -scheme UniLLMs \
  -configuration Debug \
  -sdk iphonesimulator \
  -destination 'generic/platform=iOS Simulator' \
  -derivedDataPath .DerivedData \
  -only-testing:UniLLMsTests

Dependencies

Dependencies are managed with Swift Package Manager. Locked versions are recorded in UniLLMs.xcodeproj/project.xcworkspace/xcshareddata/swiftpm/Package.resolved.

  • Yams

License

UniLLMs is licensed under the GNU Affero General Public License v3.0. The SPDX identifier is AGPL-3.0. See LICENSE for the full license text.

Copyright (C) 2026 Zayrick.

This software is distributed under AGPLv3 without any express or implied warranty. See LICENSE for the full terms.

Contact

Ver o original no GitHub

Bom para

Relacionados