GitHub - askimo-ai/.github: Askimo public

One app. Every AI model. Your files stay local.

Chat · Search your files · Run scripts · Build multi-step AI workflows · Execute AI agent skills
All offline-capable. All on your machine.

📥 Download • 📖 Docs • 💬 Discussions • ⭐ Star us on GitHub

What is Askimo?

Askimo is a native desktop AI toolkit that brings every major AI model (cloud or local) into a single, privacy-first application. It goes far beyond a chat interface - think of it as your personal AI workbench for real, end-to-end work.

🧠 Every model, one place - OpenAI, Claude, Gemini, Grok, Ollama, LM Studio, LocalAI, Docker AI, or any OpenAI-compatible endpoint. Switch in seconds.
🔒 Your data stays local - Local RAG lets you search and chat with your own files and folders without uploading anything.
⚡ Built as a real desktop app - Not a web wrapper. Starts fast, runs lean, handles thousands of messages without crashing or losing context.
🤖 Beyond chat - Run Python/Bash/JS scripts, chain multi-step AI Plans, execute AI Skills via Claude Code or Gemini CLI, and connect MCP tools - all from the same app.
🌐 Multilingual - English, 中文, 日本語, 한국어, Français, Español, Deutsch, Português, Tiếng Việt.

Our Repositories

Repo	Description
askimo	Core desktop app & CLI - Compose Multiplatform (Kotlin), GraalVM native CLI

Editions

Personal (free, open source) - everything you see in this repo. Download, self-host, modify, and contribute.

Team (paid) - adds centralized API key management so your whole team shares one set of keys without exposing them individually. Useful for companies rolling out Askimo without giving everyone their own API billing. Learn more →

Key Features

🗂 Local RAG - Index local folders, files, and URLs. Hybrid BM25 + vector retrieval with an AI classifier. Nothing leaves your machine.

📋 Plans - Chain multi-step AI pipelines from a simple form. Each step builds on the previous one. Export to PDF or Word. Define plans in YAML or generate them from plain English.

🛠 Skills - Reusable AI agents defined as Markdown files, following the Claude skills spec. Run them via Claude Code or Gemini CLI - they read/write your files, run shell commands, and stream live activity. Import skills directly from any GitHub repository.

🔌 MCP Tools - Connect any MCP-compatible server via stdio or HTTP, scoped globally or per project.

💻 Script Runner - Execute Python, Bash, and JavaScript directly from chat. Python runs in an auto-managed virtualenv with automatic dependency installation.

📊 Local Telemetry - Token usage, cost estimates, RAG performance per provider. Nothing uploaded.