Skip to content
TrustedSkillsbeta
Back to all skills

Docker Sandbox

Checked
joelhooks·vlatest·cloud

Isolate and test code in disposable Docker containers to ensure secure deployments and reproducible environments.

Repository

Install on your platform

Showing Claude Code. One command installs this skill into any of the agents below.

  1. 1

    Run this in your project

    terminal
    $ npx skills add joelhooks/joelclaw --skill docker-sandbox -a claude-code

  2. Uses the open-source skills CLI and needs Node.js. Add -g to install it for your user instead of this project.

What we know about this skill

Publisher

joelhooks

Taken from the source listing. We haven't verified who controls this account.

Install target

Live repository

Not pinned. Installing fetches whatever the repository holds at the time you run the command.

No human has reviewed this code. The checks below are a static scan of the files at one commit — nobody has run this skill or judged whether it works. A skill runs with whatever access you give your agent, so read the source before you install it.

Automated safety checks

scanned

This skill passed all 6 checks. We read its files at one commit and looked for the things that make a skill dangerous to run — not for whether it works.

  • Declares what it does: passed

    SKILL.md parses and declares a name and a description.

    SKILL.md exists, its frontmatter parses, and it declares a name and a description.

  • No undeclared network calls: passed

    Makes no network calls.

    Every host the skill dials — from its scripts, its config endpoints or the commands it tells the agent to run — is either well-known package and API infrastructure or declared in its own frontmatter. Links in documentation are not counted.

  • No obfuscated payloads: passed

    No encoded or obfuscated payloads.

    No base64, hex or compressed blob is decoded and then executed, and no encoded shell commands.

  • No credential access: passed

    Reads no keys, tokens, keychains or credential files.

    Neither the code nor the SKILL.md instructions read SSH keys, cloud credentials, keychains, token stores, browser cookie databases or dotfiles that hold secrets.

  • No pipe-to-shell installers: passed

    No pipe-to-shell or download-and-run installers.

    No `curl … | sh` pattern, in the code or in the instructions, that runs code fetched at run time.

  • Pinned to a commit: passed

    Scanned joelhooks/joelclaw at acb5544732.

    The exact upstream commit SHA that was scanned is recorded, so the result refers to specific bytes.

Commit
acb5544732
Manifest
skills/docker-sandbox/SKILL.md
Scanned
2 files, 25 lines
This is a static scan, not a review. It reads the skill's files; it does not run them, and it cannot tell you whether the skill is any good or whether the code does what its description says. A skill runs with whatever access you give your agent.
What each check looks for →

About this skill

What it does

This skill enables AI agents to execute tools like Claude and Codex within isolated Docker containers using Docker Desktop's docker sandbox feature. It provides a secure and reproducible environment for running agent tools, leveraging existing subscriptions to Claude Max and ChatGPT Pro without incurring additional API key billing. The sandboxes ensure that tool execution is contained and doesn’t impact the host system.

When to use it

  • Secure Tool Execution: When you need to run potentially risky or untrusted code within an agent tool environment.
  • Reproducible Environments: To guarantee consistent results across different runs by isolating dependencies.
  • Claude and Codex Integration: Specifically for using Claude and Codex tools, as this skill manages their authentication.
  • Automated Agent Loops: To create pre-warmed Docker sandboxes at the start of an agent loop and reuse them for multiple stories before destroying them at the end.

Key capabilities

  • Isolated execution of Claude and Codex using Docker containers.
  • Automatic management of Claude setup tokens (OAuth).
  • Integration with ChatGPT Pro authentication via ~/.codex/auth.json.
  • Support for pre-warming sandboxes to reduce execution overhead within agent loops.
  • Leverages existing Claude Max and ChatGPT Pro subscriptions.

Example prompts

  • "Create a Docker sandbox named 'my-sandbox' and run the claude tool on my project located at /path/to/project."
  • "Execute the codex tool inside an existing sandbox called 'my-sandbox', setting up authentication from a stored secret."
  • "List all currently running Docker sandboxes."

Tips & gotchas

  • Prerequisites: Requires Docker Desktop (OrbStack works) and docker sandbox version 0.11.0 or higher to be installed and configured.
  • Authentication Setup: Initial authentication for Claude requires interactive browser access, while Codex uses a local login process that generates an auth.json file. These credentials must be stored as secrets within the agent environment.
  • Token Refresh: Claude tokens expire after one year and need to be refreshed manually; Codex authentication may require re-login if it fails.