skills.rest
tomgun
tomgunCOMMUNITY
Explore all skills by this author →
GitHub

sandbox-setup

Generate Docker sandbox configuration files and firewall scripts for Claude Code.

Updated Dec 31, 2025
Search Tags:#security#github#claude#docker#firewall#sandbox#devcontainer
One-click install
npx skills add https://github.com/tomgun/agentic-framework --skill sandbox-setup-tomgun
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sandbox-setup
Source: https://github.com/tomgun/agentic-framework/tree/main/.claude/skills/sandbox-setup
Command: npx skills add https://github.com/tomgun/agentic-framework --skill sandbox-setup-tomgun

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Setting up a secure, reproducible Docker environment for running Claude Code autonomously requires many manual steps and security decisions; this Skill streamlines the process by interviewing the user, generating configuration files, and enforcing configurable guardrails.

Core Features & Use Cases

  • Interactive interview to select a security profile (Open, Standard, Locked) and drill into GitHub token and networking preferences.
  • Automated generation of .devcontainer configuration, Dockerfile, init-firewall.sh, run.sh, and optional Tier 2 settings.json for scoped permissions.
  • Guided GitHub token setup and verification, plus runtime checks that build and test the container, verify CLI auth, and confirm firewall behavior.
  • Use Case: Prepare an isolated development sandbox for autonomous code generation on a shared machine while minimizing network and repo access.

Quick Start

Ask the assistant to "set up a Docker-based Claude Code sandbox with Standard security, generate the devcontainer and firewall scripts, guide GitHub token creation, and run the verification checklist."

Frequently Asked Questions about sandbox-setup

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up a Docker sandbox for autonomous Claude Code?▼

Setting up a Docker sandbox for autonomous Claude Code involves an interactive interview that selects a security profile and generates the required .devcontainer configuration, Dockerfile, and firewall scripts automatically.

What security profiles are available for isolating an AI agent in a devcontainer?▼

Available security profiles for isolating an AI agent in a devcontainer are Open, Standard, and Locked, allowing configurable network and repo access restrictions for your environment.

How do I generate firewall rules and network allowlists for a devcontainer?▼

Generating firewall rules and network allowlists for a devcontainer requires selecting a security profile during the setup interview, which produces an init-firewall.sh script to enforce network isolation.

Do I need GitHub tokens and gh CLI authentication to run Claude Code in a Docker sandbox?▼

Running Claude Code in a Docker sandbox requires optional gh CLI authentication and repo-scoped GitHub tokens, ensuring the autonomous agent has guided and permission-scoped access to your repositories.

What Docker capabilities are required to enforce network isolation in a devcontainer?▼

Enforcing network isolation in a devcontainer requires the container to have NET_ADMIN capabilities, allowing the generated init-firewall.sh script to apply firewall rules and verify network behavior.

How can I verify that my Docker sandbox configuration is working correctly?▼

Verifying your Docker sandbox configuration is working correctly involves running built-in verification steps that build and test the container, confirm CLI authentication, and check firewall behavior.

Related Skills

Explore More Software Engineering Skills→
KarlinskySKarlinskyS

github-auth

Configure GitHub authentication via HTTPS tokens, SSH keys, or gh CLI login.

Community
Intermediate
infraspecdevinfraspecdev
3

devcontainer-init

Scaffold a secure per-repository devcontainer with pinned features and firewall allowlists.

Official
Advanced
krzemienskikrzemienski
2

skill-discovery

Scan project and user directories for SKILL.md files and generate a structured skill catalog.

Community
Intermediate
me2Docme2Doc
6

healthcheck

Audit host security and configure hardening for OpenClaw deployments.

Community
Advanced