STUzhy Py Execute Mcp
Run agent-generated Python in a remote secure sandbox with inline dependency declarations—no local venv or pip install on the developer machine.
Overview
STUzhy-py_execute_mcp is a Build-phase MCP server that runs Python in a secure remote sandbox with inline dependencies for AI coding agents.
What is this MCP server?
- Executes Python in a secure sandbox without local Python setup
- Supports declaring dependencies inline for one-shot scripts
- Smithery streamable-http remote with required Bearer Smithery API key
- Schema dated MCP 2025-09-29 in published metadata
- Version 1.0.0 Smithery package @STUzhy/py_execute_mcp
- Server schema: MCP 2025-09-29
- Package version: 1.0.0
- Auth: required Bearer Smithery API key on remote
What problem does it solve?
Letting agents run Python locally is messy, risky, and slow when every snippet needs a new venv and manual pip installs.
Who is it for?
Indie builders who want quick data checks, micro-prototypes, and dependency-light Python trials from Claude Code or Cursor without maintaining a execution environment.
Skip if: Production batch pipelines, heavy GPU workloads, or teams that forbid sending code to third-party sandboxes.
What do I get? / Deliverables
After adding the Smithery remote, your agent can execute sandboxed Python on demand and iterate on scripts using returned output in the same chat.
- Working Smithery MCP remote for remote Python execution
- Agent-issued sandbox runs with stdout/stderr feedback for iterative scripting
Recommended MCP Servers
Journey fit
Build is where you add execution backends so agents can validate scripts, prototypes, and data transforms safely. Agent-tooling is the shelf: this MCP exists to extend the agent runtime with code execution, not to ship end-user frontend or docs.
How it compares
Remote Python sandbox MCP, not a local REPL skill or full Jupyter hosting product.
Common Questions / FAQ
Who is STUzhy-py_execute_mcp for?
It is for solo builders and agent users who need safe, no-setup Python execution from MCP-capable coding assistants.
When should I use STUzhy-py_execute_mcp?
Use it during Build when you want the agent to run and verify Python—especially with inline dependency pins—without installing packages on your machine.
How do I add STUzhy-py_execute_mcp to my agent?
Register https://server.smithery.ai/@STUzhy/py_execute_mcp/mcp as a streamable-http MCP remote and set Authorization to Bearer with your Smithery API key.