Sourcery MCP
Let your coding agent request Sourcery reviews and fix the comments and security findings itself
Monty
Run Python your agent writes without spinning up a container
About
Sub-millisecond Python sandbox for AI-generated code
Monty is an alternative to Docker or hosted sandbox services when an agent needs to run small pieces of Python (the 'code mode' pattern, where the model writes code instead of chaining tool calls). You hand it the code, some inputs and a set of host functions; the sandbox runs a Python 3.14 subset, calls back into your functions when the code uses them, and returns the output. Resource limits and snapshots are built in. It ships as a Python package, an npm package and a Rust crate, and it powers Code Mode in Pydantic AI. OSS Monty is MIT licensed; Full Monty is a commercial server version that runs the same sandbox behind a WebSocket. About 8.7k GitHub stars. Repo: https://github.com/pydantic/monty.
Features
- Under 1 ms sandbox start from a warm pool
- No filesystem, env vars or network inside the sandbox
- Host functions and mounts are the only way out
- Resource limits and snapshots
- Python, JavaScript/TypeScript and Rust bindings
- Powers Code Mode in Pydantic AI
Install
Related skills
All skillsRelated articles
All articlesAI Agent Sandboxes Compared (2026)
I compare E2B, Perplexity SPACE, NVIDIA OpenShell, Anthropic srt and Kernel on isolation, pricing, free tiers, MCP support and real install commands.
Grok Bot vs Claude Code and Codex: What Each One Is For
How Grok Bot, a persistent personal agent, differs from coding agents like Claude Code and Codex, and when it makes sense to use each one or both together.
How to Give Your AI Agent Email, Memory and a Sandbox
Six building blocks that turn a chat model into an agent that can work on its own: an inbox, memory, a safe place to run code, app access, background jobs and tracing.
