This is an early release preview. You may encounter bugs.
Grok Build logo
Unclaimed

Tool coding interface

Grok Build

SpaceXAI's coding agent harness and TUI. Fullscreen, mouse interactive, extensible.

B 68/100 GitHub score ? This grade is derived from GitHub signals, not user votes. Open for the full breakdown.
Discussion

01 / Summary

About Grok Build.

Grok Build is a terminal coding agent from SpaceXAI, distributed as the grok command. It runs as a full-screen text user interface that reads the codebase, edits files, executes shell commands, searches the web, and manages long-running tasks. The same agent runs headlessly for scripting and continuous integration, and embeds in editors through the Agent Client Protocol.

The repository holds the Rust source for the command-line interface, the terminal interface, and the agent runtime, synced periodically from the SpaceXAI monorepo; a file at the root records the monorepo commit the tree corresponds to. The crates split the terminal interface (scrollback, prompt, modals, rendering) from the agent runtime and its leader, stdio, and headless entry points, the tool implementations for terminal, file editing, and search, and a workspace crate covering the host filesystem, version control, execution, and checkpoints. The workspace manifest at the root is generated and treated as read-only.

Prebuilt binaries exist for macOS, Linux, and Windows; building from source pins the Rust toolchain and resolves hermetic tools, including the protocol-buffer compiler, through DotSlash. macOS and Linux are supported build hosts, while Windows builds are best-effort and untested from this tree. The binary authenticates through the browser on first launch. External contributions are not accepted.

  • Full-screen terminal interface: scrollback, prompt, modals, and rendering, with mouse interaction and theming
  • Codebase editing and execution: file edits, shell command execution, search, and web search through built-in tools
  • Long-running task management: the agent runtime tracks work that spans more than one turn, with workspace checkpoints and version-control awareness
  • Three run modes: interactive, headless for scripting and continuous integration, and editor-embedded through the Agent Client Protocol
  • Extensibility: MCP servers, skills, plugins, and hooks, documented alongside slash commands, keyboard shortcuts, configuration, and sandboxing in the shipped user guide
  • Sandboxing: a dedicated sandbox crate in the command-line crate closure, covered by the user guide

02 / Discussion CREDIBILITY-GATED

Discussion

Reading is open to everyone. Posting and voting need a verified identity or a GitHub grade of B or higher.

  • No discussions yet.

04 / Build

Build with Grok Build.

Browse the catalog for frameworks, tools, and harnesses, each scored on real GitHub credibility.

Get Grok Build →

Browse the catalog