grok-build  by xai-org

AI coding agent harness and TUI

Created 3 weeks ago

New!

21,628 stars

Top 2.5% on SourcePulse

GitHubView on GitHub
Project Summary

Grok Build is SpaceXAI's terminal-based AI coding agent, designed to enhance developer productivity. It offers a full-screen, interactive TUI that understands codebases, facilitates file edits, executes shell commands, searches the web, and manages tasks. The tool benefits users by providing an integrated AI assistant for interactive development and automated scripting/CI workflows.

How It Works

This project provides the Rust source for the grok CLI/TUI. Its core approach involves a sophisticated terminal user interface that interacts directly with the user's codebase and environment. It leverages AI to perform actions like file manipulation and web searches, and manage complex tasks. The system supports interactive use, headless execution for scripting/CI, and integration via the Agent Client Protocol (ACP).

Quick Start & Requirements

Prebuilt binaries are available for macOS, Linux, and Windows. Installation:

  • macOS/Linux: curl -fsSL https://x.ai/cli/install.sh | bash
  • Windows PowerShell: irm https://x.ai/cli/install.ps1 | iex

Building from source requires Rust (pinned via rust-toolchain.toml) and protoc. macOS/Linux are supported build hosts; Windows builds are best-effort. To build/launch TUI from source: cargo run -p xai-grok-pager-bin. Docs: docs.x.ai/build/overview, x.ai/cli.

Highlighted Details

  • Features a fullscreen, mouse-interactive TUI for AI coding assistance.
  • Supports interactive, headless (scripting/CI), and embedded (ACP) modes.
  • Capabilities include codebase understanding, file editing, shell command execution, web searching, and task management.
  • The repository is synced periodically from the SpaceXAI monorepo.

Maintenance & Community

The project is synced periodically from the SpaceXAI monorepo. External contributions are not accepted. No community channels are mentioned.

Licensing & Compatibility

First-party code is licensed under Apache License, Version 2.0. Third-party/vendored code retains original licenses. Apache 2.0 is permissive for commercial use, but users should review vendored component licenses for restrictions.

Limitations & Caveats

Windows builds are best-effort and not actively tested from this source tree. External contributions are explicitly not accepted.

Health Check
Last Commit

1 day ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
0
Star History
21,687 stars in the last 24 days

Explore Similar Projects

Starred by Chip Huyen Chip Huyen(Author of "AI Engineering", "Designing Machine Learning Systems") and Elvis Saravia Elvis Saravia(Founder of DAIR.AI).

DeepCode by HKUDS

0.0%
16k
AI agents transform ideas into production-ready code
Created 1 year ago
Updated 5 days ago
Feedback? Help us improve.