Advai-X/advai-cli

Unified CLI for local skills, external CLIs, and terminal AI chat.

What it solves

advai-cli provides a unified command-line interface to manage AI-related tools, browser automation, and knowledge bases. It eliminates the need to switch between multiple disparate tools by offering a single entrypoint for terminal-native AI chat, local skill management for various AI agents, and browser control.

How it works

The tool is built with a Python core and can be installed via PyPI, npm, or Homebrew. It operates as a local-first system storing state in ~/.advai. Its primary components include:

  • Terminal UI (TUI): A chat interface that connects to any OpenAI-compatible API backend.
  • Browser Bridge: Uses a Chrome extension and a local daemon to execute JavaScript, take screenshots, and automate browser sessions.
  • Skills Manager: Installs and syncs "skills" (definitions/configurations) from GitHub or skills.sh into directories used by other AI agents like Cursor, Claude Code, and Windsurf.
  • OpenCLI Bridge: Integrates with the opencli ecosystem to discover, install, and execute external command-line tools.
  • Knowledge Bases: Manages local collections of markdown documents (OKF-friendly) for search and retrieval.

Who it’s for

Developers and AI power users who want a centralized way to manage their AI agent configurations, automate web tasks from the terminal, and interact with LLMs through a streamlined TUI.

Highlights

  • Unified Entrypoint: One command (advai) for chat, browser automation, and tool management.
  • Cross-Platform Agent Sync: Syncs skills across 50+ built-in platforms including Cursor, Trae, and GitHub Copilot.
  • Browser Automation: Local-first browser control (open, click, type, extract) via a Chrome extension.
  • OpenAI Compatibility: TUI supports any backend that follows the /chat/completions standard.
  • Local Knowledge Management: Create and search local knowledge bases using markdown files.

Related

  • Project
  • Project
  • Project
  • Project
  • Project