AAswordman/Operit

What it solves

Operit AI is an open-source AI Agent platform for Android that transforms a mobile device into a fully functional AI-powered workstation. It bridges the gap between LLMs and the Android system, allowing AI to perform real-world tasks like file processing, code development, and device automation instead of just chatting.

How it works

Operit connects LLMs (cloud-based or local) to the Android OS, a built-in browser, and a development environment. It uses tool calling, workflows, and extension systems (ToolPkg, MCP, Skill) to interact with the device. For automation, it leverages accessibility services, Shizuku (ADB-level permissions), and Root access to control the UI. It also includes a built-in Ubuntu 24.04 user space for terminal-based development.

Who it’s for

• Android power users who want an AI agent to automate their mobile device.
• Mobile developers who need a portable AI-assisted coding environment with a terminal and project workspaces.
• AI enthusiasts who want to run local LLMs (via MNN or llama.cpp) directly on their phones.

Highlights

• Deep Android Integration: Automates app interfaces and system capabilities using accessibility and ADB/Root permissions.
• Integrated Development Environment: Features a project workspace with code editing, syntax highlighting, and a built-in Ubuntu 24.04 terminal.
• Flexible Model Support: Supports cloud providers (OpenAI, Anthropic, Gemini) and local inference (MNN, GGUF via llama.cpp, Ollama).
• Extensible Ecosystem: Includes a unified market for scripts, ToolPkgs, and MCP servers to expand agent capabilities.
• Advanced Memory & Roles: Manages graph-based long-term memory and supports complex multi-character role-play and collaboration.