Machine

Machine is an open-source agent harness and development environment orchestrator created by katspaugh. It is designed to provision and manage isolated, reproducible virtual machine environments for running autonomous AI coding agents without exposing host credentials or polluting local workspaces. In daily operations, developers configure project workspaces using global definitions or repo-local `.machine.toml` files. Running commands such as `machine up` or the interactive `machine create` wizard sets up and boots a sandboxed VM with pre-configured repositories, profiles, shell environments, and agent forwarding. AI agents run directly within this guest environment with complete freedom to execute build and test commands. What distinguishes Machine from standard container wrappers or bare sandboxes is its focus on provisioned, ready-to-work developer environments paired with host secret isolation. Machine ensures that sensitive authentication credentials and signing keys do not live directly inside the VM, relying instead on forwarding mechanisms to maintain isolation while enabling autonomous operations. Machine is distributed as an open-source tool with Homebrew formula support, targeting macOS environments running modern Python runtimes.

Tags: agent harness, cli, coding agent, developer tools, OpenSource

Visit Machine

Key features

  • Isolated VM environments — Spins up dedicated guest sandboxes per project to safely contain autonomous agent execution. - Zero-leak secret protection — Keeps signing keys and sensitive credentials on the host without storing them inside the VM. - Interactive project wizard — Provides machine create to easily configure repositories, tool profiles, shells, and agent forwarding. - Flexible configuration — Supports both global projects.toml management and repo-local .machine.toml files. - Agent forwarding — Bridges SSH and agent forwarding into the sandbox for seamless remote repository access. - Homebrew formula distribution — Integrates with standard macOS package management for straightforward installation and updates.

Use cases

Use Case: Sandboxed Autonomous Agent Execution Run untrusted or autonomous coding agents against a local codebase inside a sandboxed VM, preventing accidental host filesystem mutations or unauthorized command execution.

Use Case: Secure Code Signing and Repository Access Allow AI agents to pull code, run CI tasks, and prepare commits without baking private SSH keys or code signing keys into the guest environment.

Use Case: Disposable Per-Project Dev Environments Quickly spin up clean, throwaway scratch VMs or project-specific runtime profiles using machine up for testing dependencies and agent workflows in complete isolation.

Who it is for

Machine is designed for macOS developers, security engineers, and teams orchestrating autonomous CLI coding agents who need isolated, disposable VM environments. It is not intended for developers seeking a standalone LLM interface or those operating exclusively in cloud-based container clusters.

Install guide

Installation Install Machine using Homebrew or from the source repository: bash # Install via Homebrew tap brew install katspaugh/machine/machine # Or clone and set up locally with Python 3.13+ git clone https://github.com/katspaugh/machine.git cd machine

First Run Run the configuration wizard to create your first project definition: bash # Launch the interactive creation wizard machine create my-project # Alternatively, launch a default scratch environment machine up default

Example Session Configure a repo-local .machine.toml and start the isolated environment: bash # In your project root, boot the machine machine up # Run commands or dispatch your coding agent inside the isolated guest environment machine exec -- claude-code "Run tests and fix failing assertions"

Part of CLI Coding Agents

Categories: AI Coding Agents, Code Assistants, Developer Tools

Related CLI agents

  • SWE-agent — SWE-agent is an open-source autonomous software engineering agent harness built and maintained by researchers from Princeton University and Stanford University.…
  • Weaver — Weaver is an open-source agent harness and orchestration framework maintained by sean35mm on GitHub. Built on top of the Bun…
  • Hivelore — Hivelore is an open-source agent harness and contextual memory orchestrator created by Doucs91. It provides persistent project memory, decision tracking,…
  • Context Bridge — Context Bridge is an open-source CLI agent harness and context orchestration utility developed by serdardb. It is built to facilitate…
  • Vibe Kanban — Vibe Kanban, originally developed by Bloop AI, is an open-source orchestration harness and visual task board designed to manage and…
  • Parallel Harness Pets — Parallel Harness Pets (parallel-harness-pets) is an open-source companion tool and harness plugin created by TevvvB. It is designed for developers…
  • AI Tools
  • Categories
  • Industries
  • CLI Coding Agents
  • MCP Servers
  • MCP Categories