aerovato/container

Persistent Linux workspaces for agentic development.

What it solves

It provides persistent, isolated Linux workspaces for software development, ensuring that coding agents and development tools are confined to a specific project environment. This prevents agents from accessing the rest of the host system and ensures a consistent environment across Windows, macOS, and Linux.

How it works

The tool uses Docker or Podman to create project-specific containers. Each project gets its own environment where packages and configurations persist across sessions. Users can customize their environment via a user-layer Dockerfile or a settings JSON file, and the project directory is mounted into the container for read-write access.

Who it’s for

Developers who use AI coding agents and want a secure, reproducible, and customizable local development environment that isolates project dependencies and agent access.

Highlights

  • Cross-platform support: Works on Windows, macOS, and Linux.
  • Persistence: Workspaces, installed packages, and configurations are saved between sessions.
  • Isolation: Agents are restricted to the project workspace, protecting the host system.
  • Customizability: Supports custom Dockerfiles, runtime flags, and mount configurations.
  • Agent Integration: Includes a host-side skill that allows AI agents to configure the tool's settings and packages.

Related

  • Dispatch
    Securing AI Agents with Isolated Docker Sandboxes

    An exploration of agent-sandbox, a tool designed to run AI coding agents within isolated Docker containers to prevent unauthorized system access and ensure execution safety.

  • Dispatch
    Docker Sandboxes for AI Agents

    Docker Sandboxes provide isolated microVM environments for AI coding agents to execute commands and install packages safely without risking the host system.

  • Project
    tastyeffectco/sandboxd

    An open-source, self-hosted AI app builder engine that uses coding agents to create and host real applications in isolated Docker sandboxes on your own server.

  • Project
    finbarr/yolobox

    A containerized sandbox for AI coding agents that allows them to run commands with sudo permissions while protecting the host home directory from accidental destruction.

  • Project
    opensandbox-group/OpenSandbox

    OpenSandbox provides isolated, secure environments on your own infrastructure for AI agents to execute code, run commands, and operate browsers or desktops.