MCP y llamadas a herramientas

Askimo

askimo-ai/askimo

A native desktop AI client for chat, local RAG, multi-step AI workflows (Plans), and agent skills, supporting multiple cloud and local LLM providers while keeping user files strictly on the machine.

Abrir repositorio oficial ↗ Página de inicio del proyecto ↗
★ 334Estrellas
⑂ 71Bifurcaciones
20Problemas abiertos
KotlinIdioma
AGPL-3.0Licencia
Q@project.QualityScorePuntuación editorial

Capturas de pantalla del proyecto

Captura de pantalla de Askimo Captura de pantalla de Askimo Captura de pantalla de Askimo

Resumen

Askimo is a Kotlin-based native desktop application designed to consolidate AI interactions into a single interface. It allows users to switch seamlessly between various cloud and local LLM providers per session. Beyond standard chat, it features local Retrieval-Augmented Generation (RAG) for querying private documents without uploading them, a 'Plans' engine for chaining multi-step AI reasoning tasks, and a 'Skills' system that delegates executable tasks to local agent runtimes like Claude Code or Gemini CLI. It also integrates Model Context Protocol (MCP) tools and includes a built-in script runner for Python, Bash, and JavaScript.

Características principales

  • Multi-provider support (OpenAI, Claude, Gemini, Grok, Ollama, LM Studio, LocalAI, Docker AI, OpenAI-compatible endpoints)
  • Local RAG with hybrid BM25 + vector retrieval and smart AI classifier
  • Agentic workflows via Plans (multi-step reasoning chains) and Skills (local agent runtimes)
  • MCP tool integration via stdio or HTTP
  • Built-in script runner for Python, Bash, and JavaScript with auto-managed virtualenv
  • Native desktop application for macOS, Windows, and Linux
  • Persistent sessions stored in local SQLite
  • Local telemetry for token usage, cost estimates, and RAG performance
  • Multi-backend web search (DuckDuckGo, Brave, Tavily, SearxNG)
  • Internationalization support for 9 languages

Requisitos, instalación y guía rápida

Download the native installer for your operating system from https://askimo.chat/download/ or the GitHub releases page at https://github.com/askimo-ai/askimo/releases. To build from source: clone the repository using `git clone https://github.com/askimo-ai/askimo.git`, navigate into the directory, and run `./gradlew :desktop:run` to launch the app, or `./gradlew :desktop:package` to build native installers.

Uso

1. Install and open Askimo. 2. Add an AI provider by pasting an API key (for cloud providers like OpenAI, Claude, Gemini) or pointing the app to a running local instance (like Ollama). 3. Start chatting, or use advanced features: index local folders for RAG, define multi-step Plans in YAML or plain English, or execute agent Skills.

Compatibilidad de modelos y casos de uso

Cloud: OpenAI, Anthropic Claude, Google Gemini, xAI Grok. Local: Ollama, LM Studio, LocalAI, Docker AI. Custom: Any OpenAI-compatible endpoint via custom base URL. VertexAI, vLLM, and Nvidia NIM are also listed as supported in the repository metadata.

Notas sobre la licencia y los riesgos

Licensed under AGPL-3.0. See https://api.github.com/licenses/agpl-3.0 and the LICENSE file in the repository for details.

Editorial verification 2026-08-02: repository URL, owner, description, license and repository statistics were reviewed. License metadata: AGPL-3.0. README was fetched for the channel draft; re-check repository dependencies, releases and model terms before production use.

Lanzamiento y mantenimiento

Not stated in the repository metadata

MemPalace

mempalace/mempalace

★ 58KPython

QwenPaw

agentscope-ai/qwenpaw

★ 33,9KPython

Composio

composiohq/composio

★ 29,5KTypeScript