GenericAgent

GenericAgent runs tasks on your computer using cloud models to decide what to do. It's for people who want an AI agent to work across websites, files and desktop apps, then save successful workflows for reuse. Model calls require API keys and an internet connection, with requests going to the selected provider. It supports Claude, Gemini, Kimi and MiniMax.

Its main distinction is how it retains task experience. After completing a task, it saves the execution approach as a reusable skill. Memory also holds lasting facts and records of earlier sessions, so later tasks can draw on work the agent has already done.

Browser automation uses TMWebdriver with a persistent Chrome or Chromium session, preserving existing logins and cookies. The agent can execute code and edit files, with additional capabilities for keyboard and mouse control, screen vision, OCR and mobile interaction through ADB. Browser and computer-control capabilities require additional setup.

The agent can extend its abilities by writing scripts, installing Python packages or calling external APIs, then retain those methods for future tasks. A terminal interface supports concurrent sessions and streaming responses; a Streamlit interface and messaging frontends such as Telegram, Discord and Lark provide other ways to interact.

The Python project uses the MIT license and supports Windows, macOS and Linux. Its bundled automation playbooks are written in Chinese and were developed on Windows, with adaptation to macOS and Linux equivalents.