Gradion Ai Ipybox
gradion-ai-ipybox is a Claude Code plugin for the Build phase that provides a Docker sandbox to run Python, shell, and MCP tool calls from your agent.
by gradion-ai · github.com/gradion-ai/ipybox
Run agent-generated Python, shell, and MCP tool calls inside a Docker sandbox instead of on your laptop.
Add it to Claude Code
Install the plugin in Claude Code. One command, paste-ready.
/plugin install gradion-ai-ipybox@gradion-ai/ipyboxBuilt to be called by your agent
Skillselion is itself an MCP server. Your agent can pull this entry and a paste-ready install config straight from the API - no copy-paste.
Retrieve this entry with skillselion.get_details("plugin:gradion-ai/ipybox") and the paste-ready config with skillselion.get_install_config("plugin:gradion-ai/ipybox").
What it does
gradion-ai-ipybox is a Claude Code plugin bundle from gradion-ai/ipybox that gives solo builders a unified execution environment for Python, shell commands, and MCP tool calls inside containerized sandboxes. If you are shipping agent workflows that must not run arbitrary code on your host, this package is meant to be the default runtime layer: Docker-backed isolation, MCP-aware programmatic invocation, and CodeAct-friendly patterns. The catalog lists three plugins in one repo so you can pick default, containerless, or sandbox-focused variants without hunting separate installs. It fits indie developers who already use Claude Code and want MCP tools and Python to share one execution box rather than ad hoc subprocess glue. Complexity is intermediate because you need Docker and comfort wiring MCP servers. It is not a hosted cloud sandbox product on its own—it is the open plugin set that connects your agent to local or self-managed containers.
Highlights
- Unified runner for Python snippets, shell commands, and programmatic MCP tool calls
- Docker-based sandbox runtime with container isolation for untrusted agent code
- Bundle includes 3 related plugins for default, with/without container, and MCP workflows
- Targets CodeAct and MCP-heavy Claude Code setups that need repeatable execution
- Keywords span ipybox, sandboxruntime, docker, python, and mcp for discovery
Why builders use it
Letting agents run Python and shell on your machine is risky and messy when MCP tools need the same execution context.
After you add the plugin bundle, agent code and MCP calls can run in an isolated ipybox container with a consistent runtime.
At a glance
- Type - Plugin in AI Agents.
- Adoption - 0 installs, 74 stars, 0 votes.
FAQ
Who is gradion-ai-ipybox for?
Claude Code users who need isolated Python, shell, and MCP execution for agent workflows on their own machine or infra.
When should I use gradion-ai-ipybox?
When you are building or testing agent tools that execute code and you want Docker isolation instead of running directly on the host.
How do I add gradion-ai-ipybox to my agent?
Install the gradion-ai/ipybox Claude Code plugin bundle from the repo, enable Docker, then register the ipybox plugins and point MCP or CodeAct flows at the sandbox runtime.
Comments
Share how you use gradion-ai-ipybox, gotchas, or tips for other indie builders.
No comments yet - be the first to share how you use it.